WORLD BANK TECHNICAL PAPER NUMBER 72 WATER SUPPLY OPERATIONS MANAGEMENT SERIES WTP72 Sept. 1987 The Reduction and Control of Unaccounted-for Water Working Guidelines Philip Jeffcoate and Arumukham Saravanapavan r ~~~~~~~~~~~~~~~~~~~~~~ > ,. - ~ ~ ~ ~ RECENT WORLD BANK TECHNICAL PAPERS No. 20. Water Quality in Hydroelectric Projects: Considerations for Planning in Tropical Forest Regions No. 21. Industrial Restructuring: Issues and Experiences in Selected Developed Economies No. 22. Energy Efficiency in the Steel Industry with Emphasis on Developing Countries No. 23. The Twinning of Institutions: Its Use as a Technical Assistance Delivery System No. 24. World Sulphur Survey No. 25. Industrialization in Sub-Saharan Africa: Strategies and Performance (also in French, 25F) No. 26. Small Enterprise Development: Economic Issues from African Experience (also in French, 26F) No. 27. Farming Systems in Africa: The Great Lakes Highlands of Zaire, Rwanda, and Burundi (also in French, 27F) No. 28. Technical Assistance and Aid Agency Staff: Alternative Techniques for Greater Effectiveness No. 29. Handpumps Testing and Development: Progress Report on Field and Laboratory Testing No. 30. Recycling from Municipal Refuse: A State-of-the-Art Review and Annotated Bibliography No. 31. Remanufacturing: The Experience of the United States and Implications for Developing Countries No. 32. World Refinery Industry: Need for Restructuring No. 33. Guidelines for Calculating Financial and Economic Rates of Return for DFC Projects (also in French, 33F, and Spanish, 33S) No. 34. Energy Efficiency in the Pulp and Paper Industry with Emphasis on Developing Countries No. 35. Potential for Energy Efficiency in the Fertilizer Industry No. 36. Aguaculture: A Component of Low Cost Sanitation Technology No. 37. Municipal Waste Processing in Europe: A Status Report on Selected Materials and Energy Recovery Projects No. 38. Bulk Shipping and Terminal Logistics No. 39. Cocoa Production: Present Constraints and Priorities for Research No. 40. Irrigation Design and Management: Experience in Thailand No. 41. Fuel Peat in Developing Countries No. 42. Administrative and Operational Procedures for Programs for Sites and Services and Area Upgrading No. 43. Farming Systems Research: A Review No. 44. Animal Health Services in Sub-Saharan Africa: Alternative Approaches No. 45. The International Road Roughness Experiment: Establishing Correlation and and a Calibration Standard for Measurements No. 46. Guidelines for Conducting and Calibrating Road Roughness Measurements No. 47. Guidelines for Evaluating the Management Information Systems of Industrial Enterprises No. 48. Handpumps Testing and Development: Proceedings of a Workshop in China No. 49. Anaerobic Digestion: Principals and Practices for Biogas Systems No. 50. Investment and Finance in Agricultural Service Cooperatives (List continues on the inside back cover.) The Reduction and Control of Unaccounted-for Water Working Guidelines - ii - WATER SUPPLY OPERATIONS MANAGEMENT SERIES (Distribution Systems Management) The management of water supply infrastructure is a complex activity, involving both the operation and maintenance of existing facilities and the construction of new facilities. Water supply managers in developing countries, however, sometimes give a lower priority to operation and maintenance of old facilities than to the construction of new ones. Moreover, the management of water supply distribution systems often has a relatively low status in many water companies compared to the attention given to production facilities. As a consequence, there are large water losses caused by the premature deterioration of water mains, frequent metering failures and decreased revenues. This, in turn, can lead to economic inefficiencies caused by doubtful investment decisions. This series of technical papers is intended for waterworks managers, distribution system maintenance engineers, and those concerned with making investment decisions for the rehabilitation or replacement of water supply facilities and the construction of new works. A subseries on Distribution Systems Management, of which this is one volume, aims to provide guidance on some of the most pressing institutional, technical and social problems faced by the managers and engineers of water supply distribution system in developing countries. Two other volumes in this subseries will be published shortly: ° Corrosion Protection of Pipelines for Water and Waste Water: Guidelines o The Selection, Testing and Maintenance of Large Water Meters in Developing Countries: Guidelines Proposed future volumes in the Distribution Systems Management subseries will cover topics such as selecting, procuring, and maintaining small domestic water meters; the political, social, institutional, and organizational implications of programs to reduce water losses; detecting and regularizing illegal service connections; and the scope for privatization in operation and maintenance. Topics likely to be addressed in the main operations management series include reducing losses at source, raw water transmission and water treatment; meter reading, billing, and collection; and the characteristics of users who connect to available supply sources. All of these concerns influence the timing and scale of additional facilities. WORLD BANK TECHNICAL PAPER NUMBER 72 WATER SUPPLY OPERATIONS MANAGEMENT SERIES The Reduction and Control of Unaccounted-for Water Working Guidelines Philip Jeffcoate and Arumukham Saravanapavan The World Bank Washington, D.C. The International Bank for Reconstruction and Development/THE WORLD BANK 18 I8 H Street, N.W Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing September 1987 Technical Papers are not formal publications of the World Bank, and are circulated to encourage discussion and comment and to communicate the results of the Bank's work quickly to the development community; citation and the use of these papers should take account of their provisional character. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. Because of the informality and to present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to for-mal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The most recent World Bank publications are described in the catalog New Publications, a new edition of which is issued in the spring and fall of each year. The complete backlist of publications is shown in the annual Index of Putblications, which contains an alphabetical title list and indexes of subjects, authors, and countries and regions; it is of value principally to libraries and institutional purchasers. The latest edition of each of these is available free of charge from the Publications Sales Unit, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. Philip Jeffcoate is director of Soleminster Limited, Caterham, England, and a consultant to the World Bank. Arumukham Saravanapavan is Senior Water and Waste Engineer in the Infrastructure and Urban Development Department of the World Bank. Library of Congress Cataloging-in-Publication Data Jeffcoate, Philip, 1912- The reduction and control of unaccounted-for water. (World Bank technical paper, ISSN 0253-7494 ; 72. Water supply operations management series) 1. Water-supply engineering--Developing countries-- Handbooks, manuals, etc. 2. Waterworks7-Developing countries--Handbooks, manuals, etc. I. Saravanapavan, Arumukham, 1925- . II. Title. III. Series: World Bank technical paper ; no. 72. IV. Series: World Bank technical paper. Water supply operations management series. TD327.J44 1987 628.1 87-23006 ISBN 0-8213-0951-X - v ABSTRACT These guideLines are in eight sections and are designed for use by Bank project officers and, through them, for the managers of water authorities in developing countries. They are suitable for authorities serving populations of 200,0100 or more and may need to be abridged or simplified for use in smaller cities. Section 2, "Summary of the Guidelines," is meant for the general manager of a utility. Since it summarizes most of the text, the interested technical reader may want to proceed directly to the following sections. THESE GUIDELINES ARE NOT INTENDED TO BE ADOPTED AS RULES. Control of UFW entails the continual repetition of a series of simple, logical processes and tasks to obtain increasingly accurate, detailed data that facilitates ever-more-efficient detection of deficiencies. If there is political resolution, domestic metering procedure may be improved relatively simply. But effective reduction of leakage is likely to demand long, persistent efforts by dedicated staff over five to ten years, followed by a permanent, organized, and disciplined maintenance thereafter. Above all, it must be stressed that modern leak detection equipment only rarely offers significant shortcuts to the achievement of satisfactory conditions. - vii - CONTENTS Section 1 ISSUES 1-1 1.1 Definitions and Significance of Terms 1-1 1.2 Nonrevenue Water (NRW) 1-2 1.3 Calculating UFW 1-3 1.4 Continuity of Supply 1-3 1.5 Recognition of the Problem 1-3 1.6 Strategy for Control 1-4 Section 2 SUMMARY OF THE GUIDELINES 2-1 2.1 First Six-to-Twelve Months: Initial 2-1 Investigations and Immediate Action Program 2.2 First Two-to-Three-Year Plan: Metering and 2-5 Associated "Nonphysical" Losses 2.3 First Five-Year Plan: Leakage Detection and 2-7 Repairs and Physical Losses 2.4 Second Five-Year Plan: Review and Update 2-10 First Five-Year Plan 2.5 Operation and Maintenance in Relation to 2-11 UFW Control 2.6 New Works in Relation to UFW Control 2-12 Annex 2A Proposed Action Program: Manager's Checklist 2-15 Annex 2B Target Program Bar Chart 2-18 Annex 2C Typical Action Program and Organization Chart 2-19 Section 3 IMMEDIATE WORKS - TASK FORCE 3-1 3.1 Definition of Unaccounted-for Water 3-1 3.2 Typical Situation Facing a Bank Project Officer 3-4 3.3 An Atypical Situation 3-6 3.4 Preliminary Checks of Basic Data 3-6 3.5 Initial Analysis 3-7 3.6 Preliminary Estimate, Budget, and Financing for 3-9 the Initial Study 3.7 Organization - UFW Task Force Approach 3-9 3.8 Task Force Operations 3-11 3.9 Production 3-13 3.10 Consumption 3-18 3.11 Leak Detection and Repairs 3-21 3.12 Surveys, Maps, and Illegalities 3-23 3.13 Advice on Policy for Further Works 3-25 3.14 Strategy 3-26 - viii - Annex 3A Typical Investigation of Consumer's Meter 3-27 Accuracy and Domestic Consumption 3B Method of Locating Leaks by Sounding 3-28 3C Costs of Various Componentss of UFW Control 3-30 3D Economic Assessment Models 3-31 3E Economic Analysis of Leak Detection 3-48 3F Draft Typical Terms of Reference for Consultants 3-53 Section 4 EXTENDED ACTION FOR REDUCING NONPHYSICAL LOSSES 4-1 4.1 General 4-1 4.2 Domestic Meters 4-1 4.3 Large Meters 4-2 4.- Meter Reading Procedure 4-3 4.5 Meter Repairs and Testing 4-4 4.6 Boundaries of Districts 4-5 4.7 Review and Further Action Program 4-6 4.8 Incentives 4-6 4.9 Women Meter Readers 4-7 4.10 Meter Reading by Contract 4-8 4.11 Three-Year Program 4-9 Section 5 EXTENDED ACTION (FIRST FIVE-YEAR PLAN) FOR REDUCING PHYSICAL LOSSES 5-1 5.1 Existing Conditions 5-1 5.2 Mapping, Records, and Network Analysis 5-2 5.3 Future Training, Labor, and Equipment 5-3 5.4 Technical Assistance 5-4 5.5 Pressure Reduction 5-5 5.6 Monitoring District Flows 5-6 5.7 Minimum Night Flows 5-7 5.8 Inspectors for Valve and Hydrant Operation 5-8 5.9 Pilot Waste Meter Districts 5-9 5.10 Night Domestic Consumption 5-11 5.11 Waste 5-11 Annex 5A Training 5-14 5B District Metering -or Control of UFW 5-16 5C Step-Testing in Waste Meter Districts 5-18 5D Typical Bylaws for the Prevention of Waste or 5-31 Misuse of Water - ix - Section 6 EXTENDED ACTION (SECOND FIVE-YEAR PLAN) FOR REDUCING PHYSICAL LOSSES 6-1 6.1 Organizational Aspects 6-1 6.2 Rechecking Data 6-1 6.3 Pressure Reduction - Potential Savings in Waste 6-2 6.4 District and Waste Metering: Potential 6-2 Achievements 6.5 Service Reservoirs 6-3 6.6 Updating of Waste Detection 6-3 6.7 Public Relations 6-6 6.8 Improved Construction and Repair Works 6-6 6.9 Improved Metering 6-6 6.10 Meter Reading 6-7 6.11 Distribution System Mapping 6-7 Section 7 OPERATION AND MAINTENANCE IN RELATION TO CONTROL 7-1 OF UFW 7.1 General 7-1 7.2 Source Stations 7-2 7.3 Distribution Networks 7-2 Section 8 NEW WORKS IN RELATION TO UFW CONTROL 8-1 8.1 Design 8-1 8.2 Materials 8-2 8.3 Equipment 8-2 8.4 Training 8-3 8.5 Construction 8-3 x ACKNOWLEDGMENTS The authors are indebted to many members of the Water Supply and Urban Development Department for their encouragement and to representatives of the Project Policy Department and the regional projects divisions for their keen interest and enthusiasm and not least for their many helpful comments; all of these contributions helped to reveal the scope of the problem and to change the emphasis in the guideline procedure. Many professional friends have contributed background data and details of current practice, particularly representatives of the British Water Research Council, who provided much information and many useful references. Special thanks are due to Jozsef Buky, Sergio Calegari, John Huang, Karl Kleiner, Robert McWilliam, Peter Ware, and Guillermo Yepes, to name a few, for their comments, advice and support. Thanks are also due to Harvey Garn, who reviewed the economic aspects of the guidelines. The authors also acknowledge with gratitude permission granted by the American Waterworks Association of the U.S.A. and the National Water Council of the U.K. to reprint sections from their economic studies. - xi - PREFACE Efforts to reduce excessive unaccounted-for water (UFW) in developing countries may be hampered by several factors: lack of awareness of the practical possibilities and the potential financial and operating benefits by top management; lack of motivation at the operational level; and, particu- larly, lack of resources. These guidelines are intended to clarify issues, correct wrong impressions, and stimulate the interest of administrators and operators in making the resources available to improve reduction and control. Recent national and international publications have identified the problem of unaccounted-for water, but little has been done to relate it to the conditions in developing countries. Nor have efforts been made to show how action can best be taken by management, in logical steps, to achieve effective control even when operational problems seem insurmountable. The waterworks manager has many problems -- perhaps more political and institutional than technical -- but, first and foremost, the aim will be to provide enough water, at adequate pressure over the whole area served, to meet increasing demands. For this the manager must have the human, technical, and financial resources to meet both investment and operating costs. Typically, the project officer faces a situation in which the authority seeks to continue its traditional policy of improving the supply of water either by developing new resources or by expanding works. Data relating to the existing supply is likely to be inadequate for a realistic assessment of the supply situation; available figures on the quantity of water produced and consumed may be so inaccurate as to call for a good deal of interpretation by a seasoned professional. The different divisions within the authority frequently lack confidence in the figures produced by the others, which perpetuates ignorance of true data and obviates opportunities for improved control. Excessive UFIW is usually a clear indication of poor workmanship and lack of maintenance. EIxcuses may be offered for delayed action, but it is necessary to begin to accurately measure and assess the data to determine immediate and long-terrn strategy, which will necessarily include improved maintenance practices. The aim of these guidelines is to demonstrate one way of proceeding, step-by-step, in producing more accurate data. A clear picture of the existing situation will facilitate the development of a strategy for whatever improvements are warranted. They may be in the area of physical supply conditions or in the equitable collection of revenue necessary for the financing of improvements. If the production of accurate data demonstrates the existence of either an excessive loss of revenue from water consumed or an excessive loss of water from leakage --- or indeed both types of loss -- the guidelines go on to show how, by cost-benefit analysis, a strategy should be developed and implemented by stages to effect those improvements found to be economically justified. - xii - The Bank has prepared a narrative slide presentation on reducing UFW, which can be obtained from the Economic Development Institute (EDI). An instructor's guide and participants manual with reading exercises and case studies are also available. Interested utility managers could order these for their staff. These guidelines have been published as a single, bound volume for ease and economy of manufacture and distribution. A looseleaf format would be more useful, however, so that revised or new sections could be added as more information is developed and so that users could add their own notes as well. Each section is self-contained, and a looseleaf format would allow relevant sections to be removed and taken on mission. We recommend that users remove the spine of this volume and punch the pages to fit a looseleaf binder. Suggestions are invited for future areas of research pertaining to UFW and related topics. 1-1 SECTION 1 ISSUES 1.1 Definitions and Significance of Terms The term "unaccounted-for water" (UFW) is self-explanatory. It represents the difference between "net production" (the volume of water delivered into a network) and "consumption" (the volume of water that can be accounted for by legitimate consumption, whether metered or not). The definition is simple, but determining the true figure can be difficult. 1.1.1 Net Production The volume of water delivered into the distribution network should be accurately metered, but frequently meters are nonexistent, out of operation, or seriously inaccurate. Similar or greater inaccuracies are found with treated water flows estimated from rated pump outlets into the distribution system. (See Section 3.1.3). 1.1.2 Consumption Where all domestic, industrial, and commercial consumption is metered, the quantity of accounted-for water must include an estimate of legitimate unmetered public supplies, such as those used for firefighting, mains flushing, public standpipes, and toilets. To put it as an equation, total consumption = metered consumption + estimated consumption for all private and public purposes. Sample test metering will probably be necessary to obtain reliable estimates. Metering is essential for large government offices or military establishments, even when no charges are levied. Where domestic consumption is not metered, it has to be estimated by a sample household study, which will require a special effort. (See Section 3.1). 1.1.3 Categories of UFW- UFW calculated from the difference between production and consumption falls into two categories: (a) Water consumed but not recorded by consumer's meters or otherwise accounted for by government or other public use. This is referred to as a "nonphysical" loss and is reflected in lost revenue. It includes water consumed through illegal connections. (b) Water lost through leakage, also referred to as "physical" loss. This is a resource loss and is reflected in the cost of production. The significance of each category varies enormously in different countries and even in different water systems in the same country. 1-2 1.1.4 Wasted Water Other losses arise from wastage by the consumer. They may arise either from wasteful use or from defective fittings. In a fully metered system it is often assumed that such losses have no impact on revenue and are excluded from estimates of UFW. This assumption is not justified because the loss is frequently not recorded on the consumer's meter. In some situations a severely progressive charge for increased consumption will discourage excessive water use for wasting water, and prompt early attention to defective fittings. The tariff for wasting water, however, must be higher than the cost of repairing fittings. (See Section 3.1). Wasted water is a conservation loss and affects the cost of production. Waste, other than underground leakage, may be best controlled by the strict application of plumbing codes and bylaws. Public education and cooperation are essential to minimize losses from defective fittings and willful misuse of public standpipes or other unmetered supplies. ("Wasted water" shou-d not be confused with "waste water," which is a term identified with sewage.) 1.1.5 Raw Water Supply is the total volume of untreated water taken from a source of supply such as a river or a borehole.l/ When raw water is taken from a source, losses occur during storage, transmission, treatment, and treated water storage (apart from normal, acceptable use in the production process). Such losses can occur from leakage, overflow, or inefficient treatment plant operation. These losses are excluded from UFW but can be of equal importance and require investigation by management at the same time. (Section 3.1.3.) 1.2 Nonrevenue Water (NRW) In a fully metered system, revenue is derived from charges for the total quantity of water recorded by consumer's meters. All other water produced and not sold in this way is referred to as "nonrevenue water." It includes: (a) unmetered water supplied to government, or for other public purposes; (b) water supplied through illegal connections; (c) water lost by leakage from the authority's network; (d) water supplied to the consumer but not recorded by the meter or included in the meter reader's record. This may include relatively low rate flows from underground leaks or dripping taps in consumer's plumbing system that will not register on the / This volume of water is sometimes called "Gross Production." 1-3 meter when there is no other draw-off, even if the meter is in mint condition; (e) Water supplied to the consumer but not recorded because of defective metering. In a system that is not fully metered, where revenue is derived partly from a property or a per capita levy, the concept of nonrevenue water is not relevant. For this reason the term nonrevenue water is not used in these guidelines. 1.3 Calculating UFW UFW is usually calculated as a percentage of net production. When expressed thus, comparison of figures for different systems can be misleading, and quoted figures must be treated with caution.2/ However, since percentage figures are the standard form of expression, they are used in these guidelines. (See Section 3.1.2). 1.4 Continuity of Supply When a continuous supply of water under pressure is not available, only limited tangiblea improvements in domestic metering or reduced leakage can be achieved in the short term. Nonetheless, much can be done in the areas of checking basic data, updating maps and records, and improving overall maintenance, which will be invaluable when the availability of water improves. A start can also be made on reduction of UFW in any areas where supplies are continuous. When a continuous supply of water is avaiLable during a limited period of the year, much more can be done through a logical, phased program to identify the extent and the nature of the problem and to achieve positive results in UFW control. 1.5 Recognition of the Problem Where records show that a UFW problem exists, it has probably existed for so long that its presence is tacitly accepted by management. Control is seen as a matter for attention sometime, but not yet. (See Section 3.2). Some authorities cite UFW levels as high as 50 percent or more of all water produced; if UFW levels are above 25 percent, there is a probelem that can no longer be disregarded. Much lower figures (15 percent and less) have been recorded in particularly well-managed authorities but,in the general 21 In Europe it is becoming increasingly common to express UFW in litres per property per hour (1/prop/h) which enables comparisons to be more meaning- ful. The concept of losses per property may be considered additionally or at a future time. An interesting approach would be to use cubic meters per property per month since customer accounts are often presented on a monthly basis. 1-4 conditions that pertain in developing countries, are unlikely to be achieved in the foreseeable future. Whether action should be taken to try and reduce UFW levels is a matter for case-by-case judgment, bearing in mind that the decision must rest ultimately on economic considerations. 106 Strategy for Control Control of UFW entails the continual repetition of a series of simple, logical processes and tasks to obtain increasingly accurate detailed data that facilitate ever-more-efficient detection of deficiencies. The next section summarizes the strategy proposed in these guidelines. 2-1 SECTION 2 SUMMARY OF THE GUIDELINES NOTE: This section summarizes the guidelines and is meant for the top management personnel of the utility. Technical specialists may therefore want to proceed directly to Sections 3 to 8. These guidelines are intended to be comprehensive and may not be applicable in all cases. The order of procedure may vary in some countries. 2.1 First Six-to-Twelve Months: Initial Investigation and Immediate Action Program 2.1.1 General Manager's Checklist of Basic Data (see Section 3.4) Where a probLem appears to exist, the first step to be taken by the general manager shouldi be to measure total water production and consumption. Where there are separate districts or zones, the next step is to check the measurement of flows to each district or zone in order to locate probable problem areas. Spot checks followed as necessary by extensive tests of all large consumers' meters (including reference to historic data on the consumer's industrial output) should be undertaken concurrently. A revised estimate should then be made of- total water consumption. Next, tests should be made to discover any major hidden leaks oh trunk mains, which are rare, and for leaks from service reservoirs and tantks, which are much more common. These may be located only by isolation, which could entail tests and repairs of large valves. Losses by overflow from service reservoirs caused by uncontrolled pumping may also be important. 2.1.2 Strategy If the basic data checks described above reveal that the level of UFW is truly excessive, an overall strategy must be devised, covering: maintenance, improvement in water supply, the basis of charges, and the justification for any change in the "status quo." This must inevitably entail a rapid appraisal of potential savings in terms of revenue and water. Set against this will be the potential cost of making such reductions. A preliminary financial analysis is required to justify an immediate policy decision. (See Section 3.5.) 2-2 2.1.3 Instructions If the basic data checks indicate that the level of UFW exceeds 25 percent of net production, a preliminary investigation is justified.!/ An instruction is required from the general manager to prepare cost estimates and budget and loan applications (if appropriate) for proceeding with a preliminary study and possibly to begin selective remapping of the system. (See Section 3.6). 2.1.4 Economic Analysis The initial calculations should then be reviewed and an economic analysis undertaken; budget allocation and financial approval for a further detailed study is required at this point. The board of directors or concerned ministry should be apprised of the problem and a commitment to further action obtained. (See Annex 3D). 2.1.5 Immediate Action Program: UFW Task Force In anticipation of ongoing need for UFW control, an Immediate Action Program should be developed. One way of ensuring that effective action is taken is for the general manag2y to appoint a well-qualified personal assistant responsible for UFW.- This position should be seen as prestigious and challenging and should include additional compensation through special financing arrangements. It should be a stepping stone to promotion and might later on be rotated to provide selected senior engineers in the utility training in UFW control. The appointee must immediately be freed from all other responsibilities. She has to be responsible directly to the general manager for the initial studies, but must have the requisite stature to work in collaboration with the assistant managers responsible for finance, operations, maintenance, water production, and distribution. She should have the authority to agree upon and implement action plans with the assistant managers. She must be allocated sufficiently experienced staff to develop a separate UFW Task Force and should be given suitable assistance to undertake special studies. Alternatively, or additionally, she may be assisted by a consultant or by Bank (or other) technical assistance. The Task Force should also be provided with special financial incentives and motivation. (See Section 3.7 and 7.) 2.1.6 Production Meters The first action of the UFW Task Force must be to check the production meters (see Sections 3.9, 1.3 and 2.3.7). If installed and apparently operating correctly, they must be checked for accuracy over the full working range because errors of up to 30 percent are not unknown, even in developed countries. If not installed or not operating, the same checks can be applied. Possible methods of checking are as follows: / It is generally accepted that UFW over 15% may be considered to be excessive in some water-short countries. 2/ A one-year temporary assignment may be a good way to begin. 2-3 (a) Night test isolation and timed volume change in clear water tank or delivery tank/service reservoir compared with recorded output for each delivery pump in turn. (b) Diversion to waste through weir or other test meter. (c) Checking or installing weir or other meter in treatment plant and adjusting estimated production by volume changes in plant and clear water tank. (d) Installing on each transmission main a temporary check meter, such as pitot, with transducer for accuracy at low flows, or an insertion turbine. Both of these can be installed under pressure. If no meters exist, the installation should be made permanent and readings taken at regular intervals thereafter. (e) Checking each estimated pump output by standard pump test using closed delivery valve data to modify characteristic curves and applying the operating total manometric head. 2.1.7 Revised Estimate Calculate the revised figure for UFW based on the more accurate data for water produced. If still more than 20 percent, proceed further with investigation. 2.1.8 Budget Approval Redefine bucdget needs and obtain financing authority for the next steps. 2.1.9 Unmetered Consumption Calculate thLe total figure for accounted-for water, which must include not only the water sold by meter but also water used for public purposes (government departments, military, and so forth). (See Section 3.10.) 2.1.10 Large Meters Carry out an initial sample check for accuracy by connecting test meter or by replacement of meter or, where possibLe, by positive measurements on delivery to tanks. Results of initial tests will determine the extent of further checking justified to reassess losses due to underregistration at low flows. (Some of this work can be undertaken earlier). (See Section 2.1.1.) 2.1.11 Revised Estimate Recalculate UFW after checks on large meters and assess the potential reduction in UFW. 2-4 2.1.12 Sample Survey Concurrently with the actions outlined in Sections 2.1.4 to 2.1.9, undertake a study of actual domestic consumption by sampling procedure in different social areas (see Annex 3A). Compare the results with the metered consumption to assess the proportion of UFW attributable to underregistration of meters, nonmetering, and other losses such as illegal connections. 2.1.13 Allocation of UFW Make a sample study of the accuracy of domestic meters and, with this data, reestimate total legal consumption. Deduct the total that is accounted for in order to determine the extent of nonphysical losses. Deduct this figure from the total UFW to determine total physical losses. If either figure exceeds 15 percent, further investigation is probably cost effective. NOTE: At this stage, the losses from illegal connections have to be regarded as a form of leakage or "physical" loss. This will be corrected later (see 3.10). If the problem of illegal connections appears acute, it should be the subject of a special survey much earlier in the Immediate Action Program. (See Section 3.12.) 2.1.14 Existing Organization Review the existing arrangements for leak detection, repairs and remapping. Examine the possibilities of making a survey of illegal connections. Undertake a preliminary cost analysis to determine the extent of additional investigation to be undertaken. This could include a preliminary review of design standards and construction and material specifications. (See Sections 3.11 and 3.12.) 2.1.15 Program and Budget Review The budget should cover all actions described below, including preliminary estimates for long-term actions. (See Section 3.13.) 2.1.16 Future Action On the basis of cost/benefit analysis (see Annex 3D), the general manager has to reconsider overall strategy and obtain the necessary authority and financial approval for proceeding with such additional works as are economically justified. (See Section 3.14.) 2.1.17 Policy Decision Policy decision and implementation is now required, including confirmation of organizational changes. Decisions may include establishing the position of UFW Manager (or equivalent) and assigning him a team, remuneration and incentive pa ents, and the appointment of consultants required for special studies.- 3J UFW Manager could, after the first year or following achievement of resuLts as head of the UFW Task Force, report to the head of the Department responsible for Operation and Maintenance. 2-5 2.2 First Two-to-Three-Year Plan: Metering and Associated "Nonphysical" Losses (see Section 4) 2.2.1 Large Meters Concentrate on large meters and institute a program of (a) more frequent reading according to size; (b) regular (at least annual) testing on-site or replacement; and (c) critical examination of records and follow- up. Consider monthly reading and quarterly testing of meters 3 inches in diameter and over.(See Sections 4.1 and 4.3.) 2.2.2 Meter Reading Procedure Study general meter reading procedure and results. Undertake spot tests by management for sample study of reliability of meters and meter readings. Undertake sample survey of illegal connections in selected areas. Take steps to eliminate irregularities and anomalies. (See Sections 4.2 and 4.4.) 2.2.3 Meter Repairs and Testing Establish a program for regular (four-to-five year) meter changing and testing. Commence color coding or similar identification method. For this purpose a sample study of the meter condition is required; as is an analysis of installation problems. (See Section 4.5.) 2.2.4 Meter Standardization Examine records of meter failures and failure to register low flows. Consider standardization appropriate to the operating conditions, of type and normal domestic diameter to minimize low-flow errors. Examine a mains flushing policy and water treatment procedures to improve water quality and reduce meter failures. Alternatively, consider the use of flow limiters instead of meters. A pilot study should be made because some flow limiters are useless under conditions of low pressure. Determine policy and purchase additional or replacement meters necessary to meet future needs. A local university might be approached to undertake these tasks as a research project. NOTE: The reading, billing, and collection procedures, including costs, should at this time be reviewed by management, with technical assistance as necessary, and designed to improve data retrieval, as well as to highlight leakages on consumers' premises and reduce irregularities, underregistration, and illegalities. 2.2.5 Meter Repair Facilities Study meter repair and testing shop procedure and capacity. Amend procedure as necessary to ensure testing at very low flows. Consider incentives for improved output of testing and repair. Study output per man and future manpower requirements. Study records of meter failures and establish necessary data retrieval procedure for reducing failures. Design any necessary expansion of meter testing, repair workshop, prepare estimates, obtain budget allocation, prepare tender documents and select contractor. 2-6 2.2.6 Boundaries of Districts Examine all possible means of correlating meter consumption statistics with distribution district boundaries, including any necessary boundary changes. (See Section 4.6.) 2.2.7 Review and Further Action Program Monitor and evaluate results and budget for further actions. (See Section 4.7.) 2.2.8 Expansion of Meter Repair Expand meter testing and repair shop, recruit and train any additional employees required. 2.2.9 Incentives Examine all possible incentive schemes to encourage employee's enthusiasm for improved metering. Consider improved status bonus payments and possible rewards to meter readers for reporting defective meters, leakage, irregularities, illegal connections, and so on. This activity should be under continual review and could well begin much earlier (see 2.2.3). Remember that, because the meter reader is normally the only employee known to the public, improved public relations must begin there. (See Section 4.8.) 2.2.10 Women Meter Readers Examine the possible employment of women meter readers who are interested (a) as the major users of water, to ensure its availability at all times and at low production costs by reduction of waste, and (b) as keepers of the domestic purse, to ensure equitable payments by all consumers. There will of course be social problems and, in some countries, cultural problems which could vitiate any such proposal. (See Section 4.9.) 2.2.11 Pilot Metering Areas Select one or more possible pilot metering areas for correlation with pilot waste meter district (see Section 2.4.5). Pilot area networks must receive intensive cleaning, regular flushing, and generally improved maintenance to determine the effect on meter failures. 2.2.12 Privatization of Meter Reading Undertake, in-house or by consultant, a cost/benefit study of privatizing all meter reading (and possibly billing) through use of an incentive type of contract. The contract may or may not include meter testing and repair, as well as the survey and remapping of consumers' services, including illegal connections. (See Section 4.10.) 2-7 2.2.13 Contract for Privatization Subject to results of study, implement policy of contract metering, starting in only one district. 2.2.14 Policy for Further Action Monitoring anid evaluation of achievements. Policy decision on all further consideration of improvements to metering and any associated billing procedures. 2.3 First Five-Year Plan: Leakage Detection and Repairs and Physical Losses NOTE: This plan is designed for completion over a five-year period. For some large utilities, or when there is a resource constraint, it may take longer. The plan should be reviewed and updated annually. Where metering and leakage problems are both evident, assume that the steps to be taken to correct both will proceed concurrently. The proposed "UFW Division" should be organized accordingly. (See Section 5.) 2.3.1 Existing Conclitions for Leakage Detection and Repairs Review existing arrangements for leak detection, and procedures for reporting and repairs. Consider all possible improvements using existing employees, including more intensive visual inspection -- particularly in conditions and in periods when pressures are high (see also Annex 3.2). Determine the extent of delays in repairing reported leaks and take action to reduce any backlog including, if necessary, temporary additional contract or other labor. Review existing equipment and assess needs. Intensify repair organization as necessary to deal with extra leaks detected. This may be required only temporarily, until the backlog of leaks found disappears. (See Section 5.1.) 2.3.2 Mapping, Records, and Network Analysis Review progress on whatever remapping of the distribution network is necessary, beginning in the areas where the greatest leakage and metering problems are expected. Remapping should include finding, exposing, marking, and numbering all valves. This work may best be done in conjunction with remapping of consumers' service connections (see 2.1.3 and 2.2.12). Consider having all the remapping completed by private contractor (covering also possibility of computer mapping in the future.) This work shall be done at the same time as that suggested in Sections 2.3.6 to 2.3.11. (See also Section 5.2.) 2.3.3 Training and Equipment Arrange for additional training in leak detection of employees selected from existing leak inspectors as well as one engineer/technician suitable for providing future extended training. The objective is to form a cadre of district inspectors by selecting honest, hard working staff for training; they too must have the opportunity for career improvement and 2-8 promotion. This training may be on the job but preferably will initially include training with a foreign water authority (see also Annex 5A). Purchase of additional equipment is considered essential. Establish a stores inventory for all tools and equipment. (See Section 5.3.) 2.3.4 Technical Assistance If considered desirable, apply for and obtain a very brief preliminary study and advice from external sources such as (a) foreign water authority, (b) individual expert, or (c) Bank or other technical assistance. This study is to assess short-term needs and to advise on action program from time to time. (See Section 5.4.) 2.3.5 Pressure Reduction Make every effort to reduce all high pressures in the distribution network at all times of day and night. First reductions should be by selecting and isolating a pilot area and monitoring the effect by meter. Pressures and resulting leakage frequently increase at night and reductions in night pressures may be achieved by simply installing in-line pressure reducing valves (PRVs), without significantly reducing the daytime pressure necessary to ensure continuous supplies. A careful survey of fringe areas to zones and reconnecting of service connections often can be rewarding. Interzonal leakage must be strictly controlled. (See Section 5.5.) 2.3.6 Monitoring Net Production and District Flows Check the 24-hour variation in production for the whole system and, as it becomes increasingly possible, for separate districts (see also Annex 5B). This normally has to be done during a period when demand is low and when it may be expected that all private tanks will have been completely filled during the earlier part of the night. (See Section 5.6.) 2.3.7 Net Production Retest production meters (or estimates). Repair or install additional meters as necessary at all treated water and source pump stations, service reservoirs (inlets and outlets), and boosters. Test for leakage in (a) service reservoirs and (b) transmission mains. Begin regular monitoring of production and district flows. (See Section 5.7.) 2.3.8 Large Meter Night Consumption Take sample readings on all large consumers' meters to assess industrial and commercial night consumption for the whole system. Adjust the total night production flows to determine the net (domestic) night demand and express as a percentage of production. This figure should not exceed 25 percent (or alternatively 10 litres/property/hour). Repeat for each district. 2.3.9 District Inspectors for Valve and Hydrant Operation Select top-grade district inspectors to have the sole responsibility for all maintenance and operation of valves, including initial location and subsequent guarding of surface boxes (see 2.4.5). During the same period, 2-9 install district (or zone) meters. Ensure that districts can be isolated permanently or, if this is not operationally possible, at least temporarily for test purposes. Concentrate leak detection and repair in districts that appear to have the most leakage. Always measure and record the quantity of leakage detected and stopped. (See Sections 5.8 and 7.3.) 2.3.10 Pilot Waste Meter Districts Select one or more pilot areas (see 2.2.11) and install a waste meter making sure that all valves can be effectively closed. Undertake regular tests of 24-hour and night flows and, if possible, regular step tests (see Annex 5C). Initially test monthly for1 raining purposes, later every six months. (See Sections 5.9 and 5.10.) - 2.3.11 Production and Consumption Study the 24-hour total production of water in relation to average total metered consumption in each district. For this purpose, boundary adjustments may have to be made to meter consumption records. 2.3.12 Physical and Nonphysical UFW Reestimate the proportions of UFW due to metering deficiencies and waste. 2.3.13 Wasted Water Review problems of wasted water, both public and private, and determine action to be taken by improved education and public relations as well as by strict application of bylaws. Review existing bylaws and make changes as needed. Where bylaws do not exist, prepare legislation for approval by the government (see Annex 5D). (Also see Section 5.11.) 2.3.14 Concentration of Effort Experience shows that most leakage problems are due to innumerable small leaks on service connections. Even when domestic consumption is metered, these small leaks can often be found on the consumer's pipework. Correcting this problem is usually the responsibility of the householder, upon whom notice should be served to take action within a certain number of days. If no action is taken, faults should be rectified at the householder's expense. Repairs are a time-consuming and tedious task. Concentrate the leak detection and repairs in districts where night line tests suggest the greatest leakage losses occur. Do not expect that correlators or other modern equipment will normally solve the problem, though they may be invaluable for locating certain leaks. Monitor and evaluate for a trial one-to-two year period. Pinpoint possible areas where complete renewal of network may be the most economical solution. 2.3.15 Economic Appraisal Make an economic study of the reduction in leakage achieved thus far. If this part of the UFW (that is, excluding an estimated amount for underregistration of consumers' meters) is still greater than 12 percent, 1/~~~~~~~. -/ Since this paper was written, more advanced accoustic correlators (para. 3.11) have been shown to be much more effective in the rapid and accurate location of leaks. As a result, step-tests, with their attendant problems of extended night work, are becoming superseded. 2-10 review the procedure to determine (a) possible improvements in control without changing the basic control method; (b) additional potential savings by introducing waste meter districts or by complete replacement of all mains and services in selected streets (see also Annex 3D). 2.3.16 Policy and Budget Reexamine policy and budget. 2.4 Second Five-Year Plan: Review and Update First Five-Year Plan NOTE: Over the first five years, reliable teams for waste detection and repair will have become established. Procedures for valve maintenance and mains flushing should also have been developed. Guidelines for further action can only be suggested in general terms. 2.4.1 Organizational Aspects Review the organization that has been set up for UFW control in the light of achievements. Either confirm or make needed organizational changes, as long as changes will not restrict further efforts toward controlling UFW. (See Section 6.1.) 2.4.2 Rechecking Data Where the marginal cost of water indicates that further reduction is justified, proceed by steps, first repeating the original program of checking accuracy of production metering and accuracy of estimates of public unmetered water consumption. Examine potential savings offered by alternative methods of procedure. (See Sections 6.2 and 6.4.) 2.4.3 Pressure Reduction Examine possible further reductions in pressure including the use of telemetry-controlled PRVs. (See Section 6.3.) 2.4.4 Service Reservoirs Recheck leakage from service reservoirs and transmission mains. (See Section 6.5.) 2.4.5 Updating Waste Detection Reexamine possible improvements using additional training, equipment, labor, and incentives. If not already instigated consider the introduction of a bonus scheme based on savings in annual operating costs resulting from reductions in UFW. Introduce competitive spirit and appropriate rewards by monitoring, evaluating, and showing in chart form the achievements of each inspection group. Extend the pilot scheme for waste meter districts by stages, beginning in those Sections where excessive leakage is expected. Step test to 2-11 measure night demand and locate demand on each street. Repeat tests every six months and compare results (see Annex 5C). Alternatively, or additionally, extend district metering by subdividing the original large districts and increase the frequency of monitoring districts (see Annex 5B). (Also see Section 6.6.) 2.4.6 Public Relations Establish respect for bylaws by appointing plumbing inspectors to improve standards of consumers' internal plumbing or by such other means as appear practicable. This could include adult education, publicity, legal action, and community involvement. The private sector could be given incentives for providing low-cost and efficient plumbing services. (See Section 6.7.) 2.4.7 New Works and Repairs Pay particular attention to improving the standard of construction of new works and repairs to ensure that all work is properly inspected and may be expected to remain leak-free for years to come. 2.5 Operation and Maintenance in Relation to UFW Control The password is cleanliness. Operator training is a prerequisite for successful operation and maintenance. If problems of grit in the meters, valves not shutting off, etc. are to be minimized, it is essential that proper attention be paid to all maintenance works, particularly on main and service repairs, to ensure thorough flushing out after completion of repairs. Additional hydrants and washouts must be installed if required for this purpose. The guidelines refer to tests on production meters and alternative production estimates based on pump output. Testing of all meters and pumps is essential, annually or more frequently, as indicated by results. Protection from dust during building alterations is often overlooked and is a common cause of failure. Pressure gauges and recorders subject to shock and vibration should be installed in duplicate. They rapidly become inaccurate and replacement becomes necessary, whether for permanent installations or for portable use. Regular testing and calibration is required at least annually. All gauges permanently installed and operating should be throttled to minimize vibration. Although failures in transmission and leading mains are normally self-evident, regular inspection is desirable throughout their whole length to minimize failures. Total leakage tests of other mains is rarely possible, except for special investigations. Maintenance includes inspection, full operation, testing of effective closure and repairs to leaking glands, etc., on all valves. All air valves, PRVs, and hydrants should be inspected, tested, and repaired. 2-12 Examine possible cleaning and relining of distribution mains to obtain a clean system. In many circumstances internal cleaning and relining of mains may be desirable and economically justified, but this normally calls for the employment of specialized contractors and should be done initially on a pilot scale. Re-examine potential improvements in operation and maintenance. One way of ensuring proper maintenance is to have "district inspectors" who have the sole responsibility for all valve operations for thie repair teams. They may also undertake waste inspection while repairs are in progress in the same locality. They are also responsible for seeing that all valve boxes remain exposed during highway maintenance and after road repairs. They are responsible for all internal cleaning of mains, disinfecting and flushing after repairs and, finally, for seeing that all isolating valves remain closed and that all valves closed for repairs are reopened. (See Section 2.3.9). All valves must have marker posts fixed, and must be numbered and recorded on distribution plans. All failures and repairs must be recorded and collated in a form which readily reveals operating conditions for each street. When special pilot areas are established, preliminary flushing out of all pipelines and testing and replacement of all valves in the pilot area are required. The cost of repairs to individual small underground leaks being relatively high, every effort must be made to ensure high quality workmanship, flushing out and testing before excavations are refilled. Every effort must also be made to prevent the perpetuation of any corrosion problems, particularly in the replacement and repair of services. (See Section 7.) (See also Corrosion Protection of Pipelines for Water and Wastewater: Guidelines, World Bank Technical Paper No. 69.) 2.6 New Works in Relation to UFW Control If the problems of leakage and underregistration are not to be perpetuated, it is essential that proper attention be paid to the design and construction of all new projects, whether for initial piped supplies or for extensions of the existing system. Concurrently with the guidelines for control of UFW, every effort is required to ensure the highest quality design and construction appropriate to local conditions. Special attention should be given to the following: 2.6.1 Design Outline design must include, or make provision for: (a) simple tests on all pumps and production meters by isolation and volume measurements on clear water tanks or reservoirs and/or by discharge to waste through test meters. Installation of adequate indicators of pressures (suction and discharge), flows, amps, volts, r.p.m. and hour counters; 2-13 (b) production meters at headworks to be supplemented by meters for delivery to, and discharge from, service reservoirs; (c) ample district meters or, at least, the chambers and bypasses for later installation to be provided. For distribution extensions, smaller metered districts to be considered and allowances made in the layout. Adequate valves and hydrants to be provided for control, isolation, and flushing as well as tappings for insertion of pitometers or turbine meters; (d) where waste meter districts are justified, include chambers and bypasses, planning one meter to serve two or three districts; and (e) avoidance of high working pressures as may be appropriate. Materials Select materials that meet local conditions (such as aggressive soils or corrosive water) and comply with international or other appropriate standards (see Section 3.7). Equipment Select equipment appropriate to local conditions, with particular regard to allow for interchanging with existing equipment, minimizing stocks of spare parts, and simplifying operation, maintenance, and repairs. Suitability for purpose should take precedence over investment cost. Simple operation is preferable to the most modern sophisticated plant, especially where labor is plentiful. Responsibility for selection of equipment ultimately rests with the Authority, whether design and construction are in-house or by consultant/contractor. For this reason close collaboration with such external agencies is essential. Training of labor in construction standards is a key factor in obtaining satisfactory new works. All construction must be required to comply with approved specifications. The same rigid specifications used for major contracts are required for small local contracts; the supervision and inspection of local contracts should be made the responsibility of a qualified consultant. For pipelaying contracts, particular attention should be paid to the provision of temporary stopends in order to prevent stones and grit from entering pipes during construction. Proper standards of construction, including depth, trench pumping, alignment, bedding, jointing, refilling, testing, swabbing, disinfecting, flushing, and retesting of all pipelines are essential. Retesting at working pressure should follow the provision of services. Service connections should be bent to give flexibility and prevent excessive strain on tappings. Ferrules (corporation cocks) and, where appropriate, saddles shLould be provided. Main stop-taps with surface boxes should be fitted outside and adjacent to consumer's boundaries to give 2-14 adequate night-time access for leakage sounding. Alternatively, surface boxes must be provided over the ferrules. Complete records should be kept of all as-built work. (See Section 8.) 2-15 Annex 2A. Proposed Action Program Manager's Checklist 2.1 Initial Investigations/ Indicative Target Time Scale Immediate Works Weeks for Tasks Weeks Cumulative (6-12 Months) Min. Max. Min. Max. 2.1.1 Check basic data 2 4 2 4 2.1.2 Devise strategy 1 2 3 6 2.1.3 Give instruction for cost estimates, budget, and loan applications 1 2 4 8 2.1.4 Budget allocation and financial approval 4 8 8 16 2.1.5 Establish a task force 2 4 10 20 2.1.6 Check on production meters 4 8 14 28 2.1.7 Revise estimate of UFW 2 4 16 32 2.1.8 Obtain further budget approval 2 4 18 36 2.1.9 Check tota]L unmetered consumptioni 2.1.10 Check large meters 2.1.11 Revise esti.mate of UFW 2 4 20 40 2.1.12 Survey actual consumption (10) (20) (20) (40) 2.1.13 Allocate UF'W to physical and nonphysical losses 2.1.14 Preliminary estimates for further action 1 2 21 42 2.1.15 Review future programs 3 6 24 48 2.1.16 Approve budget for future action 2 4 26 52 2.1.17 Decide on policy TOTALS 26 52 26 52 2-16 Manager's Checklist (Continued) 2.2 Extended Action for Reducing Indicative Target Time Scale Nonphysical Losses Weeks for Tasks Weeks Cumulative (2-3 Years) Min. Max. Min. Max. 2.2.1 Read and test large meter 4 8 4 8 2.2.2 Study meter reading procedure 8 16 12 24 2.2.3 Establish meter repairs and testing program 13 26 25 50 2.2.4 Review meter standardization 4 8 29 58 2.2.5 Study meter repair shop facilities and required extensions 8 16 37 74 2.2.6 Correlate supply and metered consumption boundaries 13 26 50 100 2.2.7 Review and approve budget for further action 4 8 54 108 2.2.8 Expand meter repairs and recruit staff 20 40 74 148 2.2.9 Consider incentives 2.2.10 Consider women meter readers 2.2.11 Establish pilot metering areas 26 52 100 200 2.2.12 Review privatizing of meter reading 13 26 113 226 2.2.13 Let contract for pilot privatization 13 26 126 252 2.2.14 Decide policy on further action required 4 8 130 260 TOTALS 130 260 130 260 2-17 Manager's Checklist (Continued) Indicative Target Time Scale 2.3 Extended Action for Reducing Weeks for Tasks Weeks Cumulative Physical Losses (5 Years) Min. Max. Min. Max. 2.3.1 Review existing leak detec- tion and repairs 4 8 4 8 2.3.2 Review progress on remapping 2 4 6 12 2.3.3 Review training and equipment 8 16 14 28 2.3.4 Obtain technical assistance 8 16 22 44 2.3.5 Reduce high pressure 13 26 35 70 2.3.6 Monitor net production and district flows 26 52 61 122 2.3.7 Retest production with meters 26 52 87 174 2.3.8 Monitor large meter night consumption rates 26 52 113 226 2.3.9 Appoint district inspectors 26 52 139 278 2.3.10 Establish pilot waste water districts 26 52 165 330 2.3.11 Study production and con- sumption 26 52 191 382 2.3.12 Reestimate proportions physical and nonphysical 4 8 195 390 2.3.13 Study control of waste and take action 26 52 221 442 2.3.14 Concentrates effort 22 44 243 486 2.3.15 Make econornic appraisals 13 26 256 512 2.3.16 Reexamine policy and budget 4 8 260 520 TOTALS 260 520 260 520 NOTE: Many of the above tasks may take longer but can be undertaken con- currently to comply with the suggested program. This table and the related bar chart (Annex 2B) provide a format for the Manager or a Project Officer to propose an action program suited to the needs of a particular utility. Responsibility for the various tasks are allocated in the organization chart at Annex 2C. Annex 2B. Target Program Bar Chart I IflllI WEEKS 0 26 52 78 104 130 156 182 200 208 234 250 260 286 300 Initial Strategy/Budget _ Establish Task Force _ Approve Amended Budget Sacnple Survey Consumption Economic Study, Budget. Policy _ - Study Large Meters Study Meter Repair & Testing (Technical Assistance) Meter Standardization Study Meter Shop Correlate Supply/Meter Boundaries CIO Extend Meter Repair Shop Establish Pilot Metering Areas Pilot Contract Privatisation Meters Review Existing Waste Control Training for Waste Control (Technical Assistance) Pressure Reduction Studies Monitoring Flows Re test Production with Meters Monitor Night Consumptions Appoint District Inspectors Stuly Procuction/Consumption Act on Waste and By Laws m Concentration ot Efforts m Study Organization World Bank-30505 I ~~~~~~~~~~~~~~~~~~~~~~~~~~~~m dy Studly Dto a l nmetered Consumption Tlechn SuY sictal Annex 2C. Typical Action Program T _| and Organization Chart |Sualtegy ondget & 'rarncingg r o r `odCi Mete, Rednng Adm,nisftotibn Leak Dfect mn rieReev" Itllegadizties o Fmunane Mapping Mat Conveyors New Meters T-ptrtiarg Leak Reptirs Metep Repairs eoindo,ed [ X < 1 } Momo~~~~~~~~ridg iniia Pl |Sduc .tion Metersl Mt.edtgo Evoluat-o Revised Estimote I'ols of Prod Mee,Recrd l ~~~~~~ ~~(St,idy) eeRcod Costmg _1 |7;;4 ;71 1 | ~~~~~~~Rete.t Rese,volls | 7 | l 1 < | Equ~~~~~clipmen | BMan Meters Tech 'Ca' { Pr,vt,s,ng (Study) | i § 8 ~~~~~District Mete-ig l ltFlushmg Study (P,Iot)| Woste Si.dy , I Study Repo,, Shop I Pressure Reduction]| Design Extension | Bunoy oratn| ws~~~~~~~~B La t L Action Study X 0 0 I I t- I l | IncreOse ~~~~~~D,,I,,c l Extended Meter Improved Metermg =mlnetlo Rear/etn Psrocedure 3-1 SECTION 3 IMMEDIATE WORKS - TASK FORCE 3.1 Definition of Unaccounted-for Water It is essential first to clarify what is meant by "unaccounted-for water (UFW)." It has been defined recently as: "The difference between net production (the volume of water delivered into a network) and consumption (the volume of water that can be accounted for by legitimate consumption, whether metered or not). The definition is simple, but determining the true figure can be difficult. It includes leakage, or "physical'" loss, but also underregistration or misreading of meters and supplies through illegal connections, or "nonphysical" loss. There has been some confusion in the past as to whether, in systems where domestic supplies are metered, the volume of water accounted for should include water supplied without charge for central and local government establishments and for public purposes such as firefighting, parks, fountains, lavatories, flushing streets, and so on. All such water can be accounted for and has to be included in consumption figures. Confusion has also arisen in countries where domestic supplies are not metered. In these countries there has been a tendency to regard all UFW as water lost through waste, (that is, not through leakage only), and to disregard the nonphysical losses resulting from underregistration of bulk supply, industrial, and commercial meters, even though such losses may represent a significant percentage of the total supply. For the purposes of these guidelines, unaccounted-for water comprises the following: (a) Water consumed but not recorded by consumers' meters or otherwise accounted for by government or public use. This referred to as a "nonphysical" loss and is reflected in lost revenue. It includes water consumed through illegal connections. (b) Water lost through leakage, also referred to as physical loss. This is a resource loss and is reflected in the cost of production. Wasted water, sometimes simply called "waste," is water that is truly lost for reasons such as leakage. When domestic supplies are not metered much attention is paid to reducing waste other than leakage, whether it is wilful or not. When domestic supplies are metered, waste by the consumer is sometimes not regarded as a serious matter because it is assumed that the water is paid for. This assumption is probably not justified, because the rates of flow from waste whether from a dripping tap, an overflow, or an underground leak, may be too low to be recorded by a meter, even though the volumes are significant because of the continuity of flow. 3-2 Waste also occurs in government institutions, and should be discouraged by levying charges according to meter readings. Waste that is actually recorded on consumers' meters and sold may be acceptable when surplus water exists, but not when there is a scarcity and reduction of waste could defer new capital expenditure. Control of waste, as indeed of UFW, normally has to be justified on economic grounds. In some situations a severely progressive charge for increased consumption will discourage excessive water use and prompt early attention to defective fittings. The tariff however, must be such that it costs more to use water needlessly than to have fittings repaired. Wasted water is a conservation loss and affects the cost of production. Waste other than underground leakage may be best controlled by the strict application of plumbing codes and bylaws. Public education and cooperation are essential to minimize losses from defective fittings and willful misuse of public standpipes or other unmetered supplies. Comparison of UFW figures between systems where domestic flows are metered and those where they are not is not meaningful. In unmetered systems, leakage or physical loss is usually the predominant factor, whereas in metered systems, underregistration and metering illegalities, or nonphysical loss, may be a significant, if not a dominant, factor. Such comparisons are also meaningless because of the doubtful reliability of data in unmetered systems. The determination of the correct figure for UFW in any system may not be easy because frequently, even in so-called fully metered systems, much unmetered water is used and these quantities have to be estimated (in firefighting, for example). Where domestic water is not metered, much more of the volume of water actually consumed has to be estimated. Terms other than UFW have been used in some studies, notably "non- revenue water." Such terms are not sufficiently explicit and can be confusing when considering water put to legitimate public use but not metered, and even in regard to domestic use of water in systems where such water is not metered. 3.1.1 Nonrevenue Water (NRW) In a fully metered system, revenue is derived from payment for the total quantity of water recorded on consumers' meters. All other water produced and not sold in this way is referred to as nonrevenue water. It includes: (a) Water supplied on an unmetered basis for government, municipal, and public purposes; (b) Water supplied through illegal connections; (c) Water lost by leakage from the authority's network; (d) Water supplied to the consumer but not recorded by the meter or included in the meter reader's record. It can include relatively low rate flows from underground leaks or dripping 3-3 taps, etc., in consumer's plumbing system that will not register on the meter when there is no other draw-off, even when the meter is in mint condition; (e) Water supplied to the consumer but not recorded because the meter is defective. In a system that is not fully metered where revenue is derived partly from a property or capita levy, the concept of nonrevenue water is not relevant. 3.1.2 Calculating UFW It has become customary to calculate UFW as a percentage of the volume of water put into the system. (It is therefore with a little reluctance that a departure from that custom should be contemplated.) But the aim must surely be to achieve a meaningful formula for measurement. One needs to be able to compare data for different systems and to have a basis for judging the range of acceptable figures under various circumstances. If, for instance, underregistration of meters is calculated as a percentage of total production, the figure will be dependent on the relative volumes consumed for industry/commerce and for domestic purposes. The same is true regarding leakage and the location of industry relative to the source of supply. Examination of the percentage figure shows that wide variations may have to be accepted and that comparisons can be misleading. Experience shows that both physical and nonphysical losses are closely related to the number of properties served. Only in scattered rural areas, where UFW probl(ems are unlikely to be significant in Bank projects, are the physical losses related more closely to the length of main than to the number of properties. It is for this reason that inlthe U.K., UFW is increasingly calculated in litres per property hour.- 3.1.3 Source, Headworks, and Treatment Losses Bank project officers have frequently become aware of physical losses of water not included in UFW. These are losses that occur before the water is put into the distribution system. They include leakages and overflows of untreated water from storage reservoirs and tanks, watershed leakage and bypasses inadequately controlled, leakage from untreated water transmission lines, and excessive use of water due to inefficient operation of the treatment process. These losses are normally regarded in a separate light, their control being a matter for the maintenance staff working on problems of source of supply, raw water transmission, and water treatment plants. 1/ An interesting approach would be to calculate it in cubic meters/property/ month since many utilities bill monthly. 3-4 They may add up to a significant loss which merits close investigation and it is appropriate that attention be directed at this problem concurrently with the instigation of improved control of UFW. These losses are excluded from UFW but can be of equal importance and require investigation by management at the same time. 3.2 Typical Situation Facing a Bank Project Officer An historical review of investigations of UFW problems over the past century is given in the valuable report by J.B. Buky (see Annex). The study draws attention to the emphasis that has been placed in recently published literature on the physical losses experienced in the more developed countries. It appears that nonphysical losses may not have been a serious problem, but they were not ignored entirely. In the U.K. 50 years ago it was considered good practice to change meters at regular intervals of five years or so, but this was not a major task because there were no domestic meters. No record has been found of any study of underregistration though it is known that a close watch was kept on consumption; enquiries would be made and explanations sought for any tailing off, or indeed of any exceptional increase, in individual metered supplies. In the U.S. there are references to metering problems in various papers in the American Water Works Association Journal over the past 40 years. In developing countries, conditions have developed during recent decades -- stemming partly from a woeful neglect of maintenance and partly from a lack of public standards -- where underregistration or incorrect reading of meters and increasing numbers of illegal connections have become, in some systems, the predominant factor in UFW. The major initial difficulty, mainly because of inadequate reliable data, is to determine the true level of UFW and assess whether or not it is excessive, bearing in mind that there are many reasonably operated systems in developed countries quoting UFW levels of up to 40 percent. A project officer may face the following situations: (a) For many years the authority has been unable to provide a continuous pressurized supply over the whole area. Conditions are relatively good if a continuous supply is available for some part of the year. (b) The general manager and all senior engineers are more than fully engaged dealing with day-to-day problems related to providing a better supply. Usually there has been a major expansion of the system, possibly with squatter housing and numerous illegal connections. (c) Production figures are usually based on estimates of doubtful validity. Of the consumers' meters, if 25 percent are not operating then probably at least another 25 percent are seriously underregistering and the remainder are under- registering to a lesser extent. Meters are probably too large 3-5 for accurate measurement of low flows. All properties have huge storage tanks for use when the suppLy fails. Most of the flow to properties is very slow due to low pressure or slow filling of storage tanks through old ball valves. This alone is the cause of much underregistration. Where the flow is not naturally slow, it may be intentionally throttled by the consumer to encourage underregistration. (d) Records may not have been updated for 20 years. Maps of the distribution system are also outdated and of little value. Valves are few; most have been covered by highway resurfacing and cannot be found since marker posts do not exist. Where valves can be found they cannot be operated effectually because of stones in the gate or bent spindles -- and the glands leak. Hydrants are sparse and are rarely used or tested; mains are never flushed, so the water contains grit or rust particles; maintenance is minimal. The new mains extensions have been badly designed and constructed with no regard for recognized standards. (e) Night demand is almost as high as day demand, but that is said to be due to the nighttime filling of numerous private tanks. That may well be, but by late night all tanks should be full or the private storage would gradually be depleted. So the late night or early morning flow should be low if it were not for leakages. In these circumstances it would be remarkable if the UFW figure were not excessive. The general manager may accept that the UFW level is high but insist that, calculated as a percentage of production, it is no higher than it has been for years. The fEact that it should have reduced as the proportion of new mains and new meter installations has increased is overlooked. The first task is to try and clarify the situation. What in fact is the true extent of UFW? How much is due to metering deficiencies and illegal connections? How much is due to leakages? Often there is an inbuilt reluctance to face up to any problem. The estimated production figures are usually thought to be too high; they often are but the supply manager prefers to have it that way. Because the finance manager doesn't believe the figures, he accepts that the total of meter consumptions is probably reasonable. The distribution manager says there is no leakage because nothing shows at ground surface. Above all, the management staff are busy; their attitude is that the problem probably does exist, but in any event things have always been thus and have come to be accepted. Moreover there is a belief that curing one leak only increases the pressure elsewhere, causing another leak. A feeling of hopelessness is engendered by the lack of resources and know-how to tackle the problems. The initial step is for the general manager and, through him, the chairman of the board to make a commitment to at least investigate the questions "How much water is accounted for? "How much of it is due to leakage?" and "What is the effect on revenue and costs?" The enthusiastic 3-6 encouragement of the finance manager is often the best way to obtain this commitment (see also Section 3.4). 3.3 An Atypical Situation In another situation, the utility could be well managed with good standards of operation and maintenance presently have a reasonable level of unaccounted-for water. However, pressures may be low and a new service of supply might be on the verge of being commissioned. This could result in a future and unexpected increase of UFW brought about by the increased pressure unless maintenance is strengthened. 3.4 Preliminary Checks of Basic Data The problems arising from lack of accurate data may appear over- whelming but they may be solved by a careful study of existing data. An attempt should be made using the existing organization to obtain the required additional data described below. Commonly occurring errors arise from inaccurate estimates of production. In the unlikely event that there is a production meter in operation, it is often found to be giving grossly inaccurate readings. Otherwise, discharges are estimated assuming that pumps are operating at duty point and in pristine condition, which is rarely the case. Where a problem appears to exist, the first step to be taken by the general manager and the existing staff should be to measure total water production and consumption. Sample tests on meters and pumps by checking volume changes in suction tanks of service reservoirs should be made and, if gross errors are discovered, these tests should be extended if possible. Failing that, all pressure and suction gauges must be checked on dead weight equipment or equivalent. Following this, a closed valve test on all except displacement pumps should be made in order to discover the delivery and suction pressures. Taking the pump characteristic curve and drawing a parallel curve through the actual points for observed total manometric head at zero flow, one can estimate the pump discharge at the normal operating pressure (see also Section 3.9). From the above tests the actual production figures may be re-assessed with adjustments for seasonal variations. The next step is to reestimate the quantity of water consumed for private and public purposes and above all for any government service for which there is no charge and which may not be metered. The public purposes include lavatories, parks, street cleaning, and firefighting, as well as the authority's own use for hydrant flushing etc. Where there are separate districts or zones, the next step is to check the measurement of flows to each district or zone in order to locate probable problem areas. 3-7 A study of the night flows may indicate whether there is excessive leakage. The minimum late night or early morning demand should not exceed 20 percent of the average 24-hour day demand. Finally, spot checks should be made on bulk supply meters and other large meters. These may be checked by volume changes in private tanks or by the use of a test meter connected in series or on a hydrant discharge. Spot checks followed as necessary by extensive tests on all large consumers' meters (including reference to historic data on the consumer's indlustrial output) should be undertaken concurrently. A revised estimate should then be made of total water consumption. Next, tests should be made to discover any major hidden leaks on trunk mains, which are rare, and for leaks from service reservoirs and tanks, whiich are much more comon. These may be located only by test isolatiorn which could entail tests and repairs of large valves. Losses by overflow from service reservoirs caused by uncontrolled pumping also may be important. After these basic checks have been made, it is possible to arrive at an estimated value of UFW and have some idea whether it is high due to leakage or due to metering problems such as underregistration, unreliable reading, or illegal connections. 3.5 Initial Analysis Before embarking on a major investigation of UFW as described above, it is useful to assess the need by a thoughtful analysis of the water supply situation from known facts. First, consider potential problems of underregistration and illegal connections. It may be that one section of the community is having to bear the whole cost of supply because of the failure to collect revenue from the remainder of the community. As a result, either revenue is insufficient for maintaining a proper supply, or a subsidy from central government or other source is required, or the tariff is inordinately high. It may be that there is a high percentage of very poor people who are unable to pay for water and receive only sporadic supplies. In such cases, a subsidy may be justified on public health or equity grounds, since insuperable difficulties in collection of tariffs would be likely, even if metered supplies were provided. Such examples may be rare and, in a normal system, either full domestic metering or no domestic metering must be the policy. If full domestic metering is decided upon, it must be operated accurately and fairly to be effective. In general, all problems of underregistration or other illegalities must be overcome. Before embarking on any form of active waste detection one has to consider the extent to which it is economically justified. 3-8 Answers to the following questions are normally available and can help in the initial analysis: (a) Is there a scarcity of water sources? (b) What are annual operating costs? (c) What would that cost be if demand were halved? (d) What would that cost be if demand were doubled? There may be circumstances in which the cost of production and distribution or expansion would be less than that of reducing leakage. In such instances, nonetheless, the consequential costs of leakage must not be overlooked. Leaks can cause damage to other structures and to expensive stored goods. When pressures are negative, polluted water can enter the system. The extra cost of treating an increased flow of sewage is also a possible factor. If water is not continuously available under pressure in some areas but is continuous in others, consider what steps can be taken to reduce waste in the fortunate areas in order to extend them. If water is scarce, the elimination of leakage may produce a continuous supply for all. If future growth will require the development of additional water supply sources, improved leak control may defer huge capital expenditure as well as reduce operating costs. Experience shows that even in relatively well operated "old" systems, leakage may not be less than 20 percent and can be as high as 40 percent. Where no active leak detection is practiced half the water put into supply is probably wasted. Experience also shows that the costs of active leak detection and repair are so high that it may not be economical in some conditions to maintain a figure less than 20 percent. If active control of leakage can reduce physical losses to 25 percent and if those operating costs that are directly related to the amount of water produced are about half the total operating costs, one can quickly deduce that a potential saving of 1/8 of the total operating costs is possible without making any spectacular claim which might be difficult to achieve. (If these savings were large enough to defer capital expenditure on additional source development, the potential saving is greater.) Thus the question to be answered is "what sum is represented by 1/8 of the total operating cost?" If such a sum will more than meet the cost of improved leak detection and repair, then the improvement is financially viable and will facilitate better management. Where water is scarce, or in a situation of exceedingly high operating cost for additional water, the option of NOT introducing active leak detection may not be available. CONCLUSIONS FOLLOWING BASIC CHECKS If the basic data checks described above reveal that the level of Ui'w is truly excessive, an overall strategy must be devised covering maintenance, improvement in the water supply, the basis of charges, 3-9 and the justification for any change in the "status quo." This must inevitably entail a rapid appraisal of potential savings in terms of revenue and water. Set against this will be the potential cost of making such reductions. A preliminary economic analysis is required to justify an immediate policy decision. 3.6 Preliminary Estimate, Budget, and Financing for the Initial Study Having examined overall strategy and decided that a reduction of UFW is desirable, it will be necessary to obtain formal approval to proceed and, for that purpose, a preliminary cost estimate must be prepared covering the initial investigation to be undertaken in the first six to twelve months. For this it must be assumed that an individual will be nominated to take responsibility for the preliminary study and that sufficient suitably experienced staff will be detailed to assist. Allowance must be made in the estimate for renewing consumption meters for larger consumers, installing additional production meters, pressure gauges and recorders, pressure reducing valves, sluice valves, and simple sounding devices. It must be accepted that the work will entail a great deal of overtime, particularly during the night. An additional allowance must be included for any outside technical assistance required to advise on the details of the study. If it can be expected that additional revenue will be quickly derived from improved registration of the larger meters, the initial investigation may prove to be self-financing. The preliminary estimate of cost for setting up the organization and mlaking the initial study will depend on the organizational changes required and the additional installations proposed. The provision, installing, and testing of new meters will be based on tenders, but the testing and repiacement or refittirng of any large valves required will need particularly careful study. Training of selected employees will be costed according to tenders received. A high estimate for the initial work is normally sound practice. If the basic data checks indicate that the level of UFW exceeds 25 percent of net production, a preliminary investigation is justified. An instruction is required from the general manager to prepare cost estimates and budget and loan applications (if appropriate) for proceeding with a preliminary study and possibly to being selective in remapping of the system. The initial calculations should then be reviewed and an economic analysis undertaken; budget allocation and financial support for a further detailed study is required at this point. The Board of Directors or concerned ministry should be appraised of the problem and a commitment to further action obtained. 3.7 Organization - UFW Task Force Approach In anticipation of ongoing need for UFW control, an immediate action program should be developed. One way of ensuring that effective action is taken is for the general manager to appoint a well- 3-10 qualified personal assistant responsible for UFW.V2 This position of the task force manager should be seen as prestigious and challenging and should include additional compensation through special financing arrangements. It should be a stepping stone to promotion and might later on be rotated to provide selected senior engineers in the utility training in UFW control. The appointee must immediately be freed from all other responsibilities. He or she has to be responsible directly to the general manager for the initial studies, but must have the requisite stature to work in collaboration with the assistant managers responsible for finance, operations, maintenance, water production, and distribution. The task force manager should have the authority to agree upon and implement action plans with the assistant managers. The taks force manager must be allocated sufficiently experienced staff to develop a seperate UFW task force and should be given suitable assistance to undertake special studies. Alternatively, or additionally, the task force manager may be assisted by a consultant or by Bank (or other) technical assistance. The task force should also be provided with special financial incentives and motivation. The next step is to determine how the investigation should proceed. The initial work may be undertaken in-house in order that the knowledge and experience thereby gained is of permanent value. Some degree of outside technical assistance to support an in-house study may be necessary. The difficulty with carrying out an in-house study is that senior employees are generally too busy with other work. This difficulty has to be overcome if progress is to be made. Alternatively the utility could obtain technical assistance from a bilateral agency, or call in a firm of UFW consultants. The latter option may entail a lengthy procedure, and the terms of reference should be3yarefully drafted to ensure that local staff receive adequate training.- With the general manager's full support, the UFW task force manager should be given the facilities and authority to: (a) obtain whatever data is necessary for the preliminary study; (b) make a best estimate of each type of UFW that exists and its extent; and (c) develop a preliminary program for control. For this purpose the task force manager should be given authority to put together a UFW task force capable of undertaking all the necessary work. This will include: tests on all available meters, pressure gauges, and pumps at works; checks on the operation and output of pumps; positive leakage tests on leading mains and service reservoirs; random accuracy tests on consumers, meters, both large and small; a review of existing leak detection and repair with estimated costs. Typical tests and studies are described in other sections of these guidelines. 2/ A one-year temporary assignment is a good way to begin. 31 These guidelines could be given to the Technical Assistance Team to review for comments and with their agreement made part of their TOR. 3-11 This program may not be acceptable in certain conditions, and some alternative may have to be adopted to avoid problems with the existing personnel and establishment. It may be that the existing organization for leak detection is considered sound but that its past activities and achievements have been restricted only by lack of policy, direction, or finance. But whatever the circumstances peculiar to an authority, experience shows that effective action can be expected only if the financial manager is fully involved and if UFW control is vested in someone not only dedicated to achievement of results but also with the necessary patience and persistence to pursue a steady methodical course over an extended period. It must be recognized that results, certainly for leak detection and repair, will not be achieved quickly and that much unrewarded effort may be necessary. The employee will have to take charge of a task force and to imbue his team with the same dedication and enthusiasm, because they may all need to work unsocial hours and face many disappointments. Above all they must not be diverted from their normal duties in UFW control, except in the event of a dire emergency. Each time a waste water inspector is called upon to undertake other duties, the impetus is lost. 3.8 Task Force Operations The task force should begin by examining all relevant data collected during the Basic Data Check and carry out the following additional tests: 3.8.1 Production i. tests on production meters, pressure gauges, pumps, and operational wasted water; ii. tests on delivery mains and service reservoirs; iii. tests on zone meters, if any, and leakage between zones; iv. tests on night flows of entire districts and zones and calculation of net (domestic only) night flows by reference to 3.7; and v. tests of effect on production of reducing pressures for entire area, if practicable, and for selected areas. 3.8.2 Consumption i. tests on bulk supply meters and all large industrial or commercial meters; ii. examination of records of large meter readings and investigation of reasons for past changes; iii. tests on all large meter night consumptions; iv. sample study of accuracy of small meters, which may require technical assistance; 3-12 v. sample study of actual domestic consumption (metered, unmetered, illegal) over 24 hours; vi. calculation of nonphysicaL losses; and vii. a review of organization, output of meter readers, and other cost data as well as the meter testing/repair arrangements and costs. 3.8.3 Leak Detection and Repairs i. study existing organization, personnel, overtime, costs etc., records of leaks found, soundings per week, leaks repaired, and time delays; ii. study possibilities for improved leakage control; iii. study training requirements and assistance available; and iv. estimate costs for the reconditioning or replacement of old mains and services; v. study pipe materials, soil aggressivity, and corrosion; and vi. study construction standards and workmanship. 3.8.4 Mapping i. study existing maps, records of connections, and updating arrangements. Obtain quotations for any necessary remapping by contract and compare with alternative in-house operation; ii, survey illegal connections; and iii. study location, condition, and operation and maintenance of valves, hydrants etc. Consider appointment of district inspectors or other measures for improved operation and maintenance. 3.8.5 Advice on Policy i. prepare a cost estimate of various additional investigations required, including recruitment and training; ii. prepare a cost/benefit analysis of potential reductions in UFW; iii. develop a plan for future organization, personnel, equipment and training; and iv. consider the use of consultants/technical assistants/ specialized contractors. 3-13 3.9 Production Tests on production meters are essential because even newly installed meters have been found to have errors of up to 30 percent. Usually the production meters are found to be out of operation but, if operating, the readings should be compared with estimated pump outputs and, after checking for zero, by direct volume measurements on isolated suction tanks or service reservoirs. Where meters are not in operation, pump outputs should be checked similarly by volume changes. Where this is not possible, and cannot be achieved by simple pipework changes, characteristic curves may be used with discretion. First, all delivery and suction pressure gauges must be tested because they are frequently incorrect. Loss of output and efficiency since the installation of thie pump can be determined by reading the pressure gauges and the total manometric head with closed delivery valve (see Figure 3-1). Plotting a parallel ch.aracteristic curve through the closed valve total pressure ordinate as recorded will give the output at normal operating pressure. Operational practice must be investigated to ensure that losses do not occur because of uncontrolled pumping and resulting overflow from service reservoirs at night. Tests for leakages from trunk mains and service reservoirs must be made, again by observation of volume changes under isolation conditions. (Isolation may be difficult to ensure; closed valves must be sounded to verify that they shut off tight.) Tests on production meters, or on pumps, may be separated to indicate total production to separate zones or districts. If any district meters exist they must also be checked for accuracy. Leakage between zones must be investigated and, if found, should be stopped permanently. If this is impracticable for supply reasons, it may possibly be stopped temporarily for test purposes and actual interzone leakage may then be assessed. Where no production meters exist and tests by volume changes are impracticable, an effort must be made to measure flows by independent meters. It may be possible to measure flows in treatment plants by inserting temporary flumes or weirs. Insertion of venturi meters in large trunk mains is an expensive exercise that cannot be contemplated. Modern alternatives are electromagnetic and supersonic meters but the installation of either of these could entail shut-down of the delivery main, which might be extremely difficult while maintaining the supply. The simple versions of the supersonic meter are prone to errors arising from variations in the upstream veLocity profile. The alternative is to insert a meter under pressure without stopping normal flows. In the past the insertion meter used in this way was a pitot, but this type of meter does not record low flows accurately. If the modern transducer is used in pLace of the old manometer to measure the pressure difference, accuracy is improved. Otherwise, it is now possible to insert under pressure a small turbine meter that measures pipe velocities in the same way as the pitot and is accurate over a wide range. 3-14 FIGURE 3-1 TYPICAL PUMP CHARACTERISTIC CURVE Estimated Pump Discharge Based on Test with Closed Delivery Valve 000 ~Effic ency Curve ~~~~oeDelivery Valve Pressure On Test ___ >-<; ____ - X - - -Normal Operating Pressure -t s--4-~~~~~~Spcified Duty Head/Volume Curve a 21~~0 I ~~~~aII Quantitv Pumped Word 8ank~-30505 A 3-15 Having tested production meters and pumps, the next important information to be obtained is a measure of the rate of demand for a continuous period of many days, if possible9 or 24 hours at a minimum. The actual consumption figures may not be easy to determine because they are masked by the filling of private storage tanks. But for supply from tanks to be continuous they must fill, like service reservoirs, before daytime draw-off recommences. That is, by the early hours of the morning, all tanks should be filled and the demancd at that time will reflect only the true consumption plus loss from leaks. It is accepted that when seasonal demands are at their highest, this may not occur; tanks do not completely fill; gradualLy they become less and less full as the days go by and eventually supplies become discontinuous. It may be that conditions are so serious that supplies are always discontinuous, but this is rare and in such conditions it is not possible to measure the true nighc demand. Otherwise, true night demand figures can be obtained during periods when demands are low. To obtain the domestic or net night demand it is necessary to deduct the demand for bulk supplies, industry, commerce, hospitals, public lavatories, and other nondomestic purposes (see Section 3.8.2.iii)e The clearest indication of physical losses is reflected, without doubt, by the figure for minimum net night flow from sources, adjusted by gain or loss from service reservoirs. This figure should not be more than 20 percent of the average domestic demand. Figure 3-2 shows what can be done by active waste control. Careful investigations by some authorities have shown that th4,true night demand is no more than 15 percent of the average 24-hour demand.- This figure may be higher in other cities depending on social conditions, hours of work, shift working etc. In recent years night domestic demand in some cities has increased due to the operation of domestic washing machines at night, in order to use off-peak electricity. Excessive figures for night demand are a clear indication that large leakages are one cause of high levels of UFW. Having determined the nigat demand figure for the whole area, it may then be possible to calculate similarly for each zone or district in the system. Otherwise this will have Zo be deferred to the second five-year plan. Losses from leakage are highest at night, when pressures are high. The results of trials in a recent experimental research program has shown that the relationship between high pressure and the rate of leakage does not follow the square root and is not even a straight line. Doubling the pressure increases the leakage by about two--and-a-half to three times (see Figure 3-3). In practice, some variation to this curve may be experienced. Studies should be made on selected parts oi^ the system to determine whether operating pressures can be reduced without producing discontinuous supplies during peak demand periods. It may be found possible to introduce pressure reducing valves (PRV's) into certain leading mains, which will prevent a rise in 4/ Where domestic supplies are not metered and the whole of UFW is related to leakage, the current aim in a well organized authority is to maintain the net night demand below 5 to 10 1/prop/hour. (Where no active leak control is practiced the figure may be 30 1/prop/h or higher.) FIGURE 3-2 Example of Net District Flow Before and After Active Leak Control Distrit 1 Daffy VPM efore RPAair-Aim Daly PJxi After Repeir-4dw 15 14- 13 12 11 10 4 ~ ~ ~ ~ 3 Day I D:y 2 DJy 3 Day 4 World Boank-30505 4 3-17 FIGURE 3-3 Typical Relationship Between Leakage (Net Night Flow) and Pressure 100 90 80 70 60 o50/ a) 40 30 _ 20. 10f G) 0 0 20 40 60 80 100 Average Zone Night Pressure (m) World Bank-30505:5 3-18 operating pressures on the district at night without restricting supplies or significantly reducing daytime pressures. The effects of reduced pressures on night demand should be determined. By now there should be some early impressions on the standards for operation and maintenance, and preliminary organizational changes in the water production area could be proposed. The first action of the task force must be to recheck the production meters. If installed and apparently operating correctly they must be checked for accuracy over the full working range. If not installed or not operating, the same checks can be made. Possible methods of checking are as follows: (a) Night test isolation and timed volume change in clear water tank or delivery tank/service reservoir compared with recorded output for each delivery pump in turn. (b) Diversion to waste through weir or other test meter. (c) Checking or installing weir or other meter in treatment plant and adjusting estimated net production by volume changes in plant and clear water tank. (d) Installing on each transmission main a temporary check meter, such as pitot with transducer for accuracy at low flows, or an insertion turbine. Both of these can be installed under pressure. If no meters exist, the installation should be made permanent and readings taken at regular intervals thereafter. (e) Checking each estimated pump output by standard pump test using closed delivery valve data to modify characteristic curves and applying the operating total manometric head. Calculate the revised figure for UJFW based on the more accurate data for water produced. If still more than 20 percent proceed further with investigation. Redefine budget needs and obtain financing authority for the next steps. 3.10 Consumption At this stage a concurrent study is needed of the operation, testing, and reading of all consumers' meters beginning with the large meters used for bulk supplies and for industry, hospitals etc. Sample checks on the accuracy of these meters using test meters, or possibly by volume changes on private tanks, is normally a simple procedure. The total "large meter" demand may then be adjusted to allow for the estimated inaccuracies based on the sample study. Concentrated attention to ensuring the accuracy of the large meters can pay handsome dividends in increasing the revenue from water sales. This may be rewarding work that yields a renid return for the increased effort. 3-19 Instead of (or in addition to) checking the accuracy of all large meters, it may be helpful to examine historical data relating to the volumes recorded and charged for in the accounts of heavy users. Unexplainable variations in monthly or quarterly consumption unrelated to production figures for the particular industry warrant investigation. Such investigations may be helpful to the industries concerned in pinpointing waste or leakage on their premises. Having checked and, by repairing and replacing meters, eliminated serious inaccuracies in the large meters, arrangements should be made to determine the rates of demand through these meters during the night. These figures can then be deducted from the gross night production figures to determine net domestic night demand. Calculate the total figure for accounted-for water, which must include not only the water sold by meter but also water used for public purposes (government departments, military and so forth). Check all large meters. Carry out an initial sample check for accuracy by connecting test meters or by replacement of meter or, where possible, by positive measurements on delivery to tanks. Results of initial tests will determine the extent of further checking justified to reassess losses due to underregistration at low flows. (Some of this work can be undertaken earlier.) Recalculate UFW after checks on large meters and assess the potential reduction in IJFW. A further sample study should then be made on the accuracy of the domestic meters, but this must be undertaken with care to give meaningful results and may require technical assistance from an outside source. Of particular importance is the accuracy at low flows (see Annex 3B). Examination of the total supply provided through domestic meters is helpful. If this figure, for example, is less than 100 liters per capita per day there may be serious anomalies in the meter sales and much of the UFW may be attributable to nonphysical loss. Interpretation of this figure will vary according to social conditions. It is also helpful to compare this figure with the total domestic supply as calculated by the total production less the total bulk, industrial, and commercial supply. This survey should reveal any irregularities in the reading of meters that may exist in some cities. As a corollary to this study of domestic meters, a sample survey can be made concurrently of the actual domestic consumption as revealed by test meters (or by normal meters after testing). This has to be done for different areas where social conditions differ. True domestic consumption may then be assessed for this particular water authority, which may be more helpful than adopting a standard such as 100 liters per capita per day. Reference has been made to the importance of accuracy at low flows of domestic meters. Particular problems arise when all the supply passes through a ballvalve to a large private storage tank. When the ballvalve is new, the period of shutting (when low flows occur) is short, but when the ballvalve is old, extended closing periods occur and much of the water is taken so slowly that the meter does not register. This is serious enough, but the problem is compounded if 3-20 the storage tank is large and the inflow is throttled by the consumer to ensure that it is always very slow; underregistration then becomes continuous. Very recent research shows that in almost half the properties with tank supplies, significant underregistration may be expected. Adjustment of the tariff could recover lost revenue but is inequitable because of the large variations in underregistration. Turbine-type meters are the most susceptible to error and should be replaced by the semi-positive displacement type when possible. (If the water supply contains large amounts of grit or rust particles, this may not be practicable.) Nominal domestic flow rates of domestic meters are usually 1.5 cubic metres per hour, but recent developments suggest that meters with a nominal flow rate of only 0.75 cubic meters per hour can be used where the supply is to a storage tank. A wholesale change may well be justified by the potential increase in revenue. It is appropriate at this stage to study this possibility and to undertake a cost/benefit analysis (see also Section 4.2). Concurrently with the actions outlined in Sections 2.1.4 to 2.1.9, undertake a study of actual domestic consumption by sampling procedure in different social areas (see Annex 3A). Compare the results with the metered consumption to assess the proportion of UFW attributable to underregistration of meters, nonmetering, and other losses such as illegal connections. After the sample analysis of domestic meter accuracy has been undertaken, it will be possible to estimate the nonphysical part of the estimated UFW. The total level of UFW has been reassessed following accurate measurement of production and unmetered supplies. It is thus possible at this stage to determine how much of the total UFW is due to physical losses and how much to nonphysical losses. It also provides the information necessary for the evaluation of possible reduction of nonphysical UFW losses and the estimation of costs of further investigations and actions necessary. Make a sample study of the accuracy of domestic meters and, with this data, reestimate total legal consumption. Deduct the total that is accounted for in order to determine the extent of nonphysical losses. Deduct this figure from the total UFW to determine total physical losses. If either figure exceeds 15 percent, further investigation is probably cost effective. At this stage, the losses from illegal connections have to be regarded as a form of leakage or "physical" loss. This will be corrected later (see 3.9). If the problem of illegal connections appears acute, it should be the subject of a special survey much earlier in the immediate action program (see 3.11). Additional information is required to review the organization necessary for undertaking any further work. This will include a review of the present arrangements for the reading, changing, repair, and testing of all meters. It is particularly important to ensure that the utmost emphasis is placed on the testing of all meters at very low flows and, where accuracy at such flows cannot be achieved, to replace the meters. It is also appropriate at this time to review all aspects of reading and billing procedures. 3-21 3.11 Leak Detection and Repairs Before investigating any possible reduction in leakage, a study must be made of the existing organization and effectiveness of leak detection already being undertaken by the authority. Normally the only action taken is known as "passive." That is, there is no team of employees actively engaged in looking for leaks. Action results only from loss of pressure and supply or actual leaks observed at ground surface by employees, the police, or the general public. Reports are received and action is then taken by inspectors to trace the source of leakage before a repair team is called in, Leaks not showing at the surface probably discharge directly into sewers and drains and are not easily found. A review is required of the leakages found and the delay in repair. Where there are inspectors already engaged in locating pipes and sounding for leaks, data is requiresd showing the number of leaks mended, volume of water saved and, particularly, the personnel, equipment, and costs of this work in relation to the estimated benefit. Active leak dletection can take various forms. Given the time and organization, the ideal approach would be that described in Section 6.0 (i) to (iv) of Buky's report of June 1982., But time rarely allows, and as a rule, immediate results are sought that require minimal effort. The simplest approach would be to call in experlenced inspection teams to undertake methodical observations and sounding throughout the day or at night, listening for leaks over the whole system. Methodical observation by experienced inspectors can be very effective. Possible leaks may also be located visually from damp ground, discoloration of walls, recent excavations, crossings by other services, uneven road or pavement surface, and so on. All valves and hydrants call for special attention. Sounding is best carried out by listening on consumersl stop taps, main valves, and hydraLnts. In some systems the consumers' stop taps are covered over and not accessible, or else are on their premises and are not accessible at night. There are then only very few places where direct contact with the mains or service connections is possible at night. Sounding must then be carried out on the ground surface, but this is not normally very rewarding. Whether to undertake sounding at night or in the daytime depends on labor costs and operating conditions. Sounding in the daytime is difficult not only because there are many extraneous noises, but also because pressures are low and the sound, as well as the actual volume of leakage, is reduced. Sounding has been found less effective for most leaks when the pressure is less than about 15 meters. One major problem is that large leaks can occur without much noise being produced, whereas small leaks may make a great deal of noise. Initial training followed by long experience are invaluable identifying real problems. For most systems the precise location of the service connections and of most of the mains are not known. They are located normally by using specialized electromagnetic equipment which, with careful use, can reveal the depth as well as the line of the pipe. 3-22 More complete details of sounding problems and practices are given in Annex 3B. A modern device for leak detection is the acoustic correlator. This was invented and first developed in the U.K., but then underwent improvements in Japan and France. Further developments have continued by a leap-frogging process in all three countries, and their equipment is now comparable. The function of this equipment is to measure the difference in time taken for a sound to travel to two pickup points. This equipment is valuable for the accurate location of a leak not traceable by other means. For normal leaks, the traditional means of sounding is effective and much quicker. The acoustic correlator is invaluable at certain times, but its use has to be restricted for detection to be effected economically. The use of conventional rods by experienced personnel is often the most cost-effective method of leak detection. In developing countries where labor costs may be low, the employment of large numbers of detection teams rather than any sophisticated metering system may prove to be best in economic terms. In developing countries, a team consisting of an inspector and at least one laborer/assistant be able to make no more than 100 soundings a day. Thus, for example, to sound 100,000 properties twice a year requires 2,000 team days, or eight teams. One may estimate needs and costs for simple active control in this way, keeping in mind local problems, including the actual performance of existing inspection teams. It may be possible for one means of transport to serve at least two teams, thereby adding only 15 percent to labor costs. District Metering More effective leak detection may be achieved more quickly and economically by concentrating efforts in those areas where leakages are prevalent due to the age, material, and condition of the pipes or to ground conditions or other factors. This concentration can be logically developed by district metering. The supply network is broken down, by isolating valves, into different districts each of which can be metered to determine demand either over 24 hours or at night. Such districts occur naturally where there are different pressure zones, but these may be divided further as appears practicable. It may not be possible for districts to be isolated permanently (that is, in a ring main system), but temporary isolation and monitoring is usually feasible. It is essential to be able to show that isolation is effective. Regular tests of demand for each district will show where leakage appears to be most extensive and will also show any changes occurring due to increased leakage. The normal district will contain about 5,000 to 10,000 properties. The net night demand in each district is a crucial figure to obtain. District metering has the additional advantage of showing management how the system is working. Gradually the areas can be broken down by progressively reducing the size of each district to, for example, 2,000 properties and, in developing countries, this may be the most effective method of active waste control (see Annex 5B). 3-23 Action may be concentrated still further (if found to be justified in economic terms) by breaking down each district into separate waste meter districts, each of which can be effectively isolated by valves that shut off tight. Each such waste district can be supplied through a single recording waste water meter and, by sequential valve operation on a night test, the night demand can be recorded for each street in the district. Regular "stepped" night tests of this type show up variations in night demand normally attributable to leakage, and this determines where detailed inspection can be concentrated to produce the most effective results. Stepped testing will be found impracticable unless there is adequate pressure, at least 15 metres, during the night (see Annex 5C). The sensible first step in any improved leakage control program is to introduce district metering. In most developing countries this alone can be a mammoth task because valves are frequently inoperable and most if not all have to be changed to ensure that district isolation is achieved. Only after this has been done can waste meter districts be contemplated and then only on a pilot scale. Above all, it is essential to have a cadre of inspectors trained in a logical approach to the problem, in coping with the various difficulties encountered in the detection of leaks, and in the use of modern equipment to make the tasks less difficult. Any new program of active leak detection must begin with at least one trained inspector for each 50,000 properties. This number will probably have to be doubled or quadrupled over a period. Preferably, a few selected employees should receive ten weeks of training by a water authority in a developed country. The first trainees should normally include semi-skilled employees as well as one engineer/technician who will subsequently be able to undertake the training of additional inspectors at home. Thought should aLso be given to obtaining additional outside assistance from the country selected for training. This assistance is likely to be required for specialized advice on the initial testing, and on district meter installation and operation during the early stages of development of active waste detection. The cost for initial training in the U.K. (including accommodations and travel) for an Engineer/Trainer and two junior engineers from a Middle-Eastern country was US$25,000 in 1985. In some cities growth has been such that the existing water supply network is inadequate to provide a continuous supply. Where skyscrapers have replaced two-story buildlings, the complete replacement of all mains and services may be necessary. Extensive reinforcement may be an alternative, but if the existing pipework is very old the cost of replacement may be met by the resulting reduction in UFW. A sampLe study is required for refurbishing a typical pilot section. For this purpose the metering of demand for that section throughout a 24-hour period is required. Where in situ cement lining has been emplaced by specialized contractors, experience has frequently shown improvements in fLows and pressures as well as reductions in leakage. It all depends whether leakages are occurring mostly in the mains, in the services, or at the service connections. 3.12 Surveys, Maps, and Illegalities There exists a body of feeling that in most developed countries standards are such that little can be done in the way of improved control of 3-24 UFW until existing maps and records have been carefully checked and updated. This may be true, but in developing countries conditions are likely to be such that, although accurate mapping will be necessary before control can be very effective, much can be done to reduce UFW while mapping and updating of records is in progress. The checklist includes much that should begin and show tangible results before accurate mapping and updating of records may be initiated. Initially it will be necessary to review existing records and decide which parts of the system require remapping. A clear, detailed picture of the system must be made available both for careful upgrading of the metering of total consumption and for the final planning of separately metered zones and districts. However, in many developing countries the physical effort of locating all valves and hydrants, testing their ability to open and close, and replacing them when necessary is essential before a district or a pressure zone can be effectively isolated. The accurate location of long consumers' services near zonal boundaries is also essential before the boundaries can be delineated. Mapping must be started immediately by the task force; this will have permanent value for the operation of the system, even if extensive UFW control is not found to be economically justified. The work of locating mains and services and accurately recording them is likely to be beyond the scope of the authority's staff, which is fully occupied on other matters. For this reason it may be desirable, or in fact necessary, for this work to be undertaken by specialized contractors. Many companies specialize in such work. which may entail no more than visual inspection and location of valves but is much more likely to require that all mains, including buried valves, be electronically traced and plotted on new maps. The cost of the latter might, for budget purposes, be estimated as $1,500/$2,000 per sheet, size Al, scale 1:1,250/2,500. The cost of this work could be as high as $50,000 a month and require many months work for a large city. Illegal Connections An intractable problem in many cities is the existence of numerous illegal connections. They are difficult to trace and, after removal, are frequently replaced. It is reported that in many cities they represent a major cause of nonphysical losses. In such cases a separate mapping exercise to locate, list and map all such connections will be required. This type of work has been undertaken by students, and this may be appropriate. If not, some other means (such as an aerial survey) must be adopted. Whatever method is used, it will be essential for management to sample check at regular intervals areas that have been surveyed and mapped to ensure that illegal connections that have been removed have not subsequently been replaced. A major problem in many cities arises from lack of supervision and maintenance of the distribution system over an extended period of years. In fact, high figures for UFW are usually associated with inadequate maintenance of works. The lack of adequate maintenance is often revealed by the failures of valves to operate effectively or even by the loss of valves and hydrants. The location of all valves and hydrants is an essential part of the mapping process and this must be followed by repair, replacement, and the provision of new surface boxes and marker posts. They must also all be numbered for future reference and use in the leak detection work. 3-25 Serious consideration must be given to the appointment of senior inspectors to be solely responsible for the operation of all valves during repair work and for ensuring that valves are maintained in full working order and are not covered over in new road surfacing work. Some organization of this type must be adopted to ensure that improved standards being introduced will be perpetuated. If this is not done, remapping will prove of little value (see Section 5.8). Review the existing arrangements for leak detection, repairs and remapping. Examine the possibilities of making a survey of illegal connections. Undertake a preliminary cost analysis to determine the extent of additional investigations to be undertaken. This could include a preliminary review of design standards and construction and material specifications. 3.13 Advice on Policy for Further Works The task force should at this stage have obtained sufficient data to assess the many problems to be confronted and to determine what additional work is justified economically. An estimate of cost is now required of all the additional investigations, including the recruitment and training of additional staff required and of retaining any outside advisors, specialized contractors, or other technical assistance considered necessary to ensure the success of proposed improvements in control. Following this, a cautious assumption may be made of the potential savings in UFW both in nonphysical and physical losses, and a cost/benefit analysis is then possible based: first, on potential increases in revenue and savings in production costs; second, on reduced loan charges by deferrment of capital expenditure on new source, treatment, and transmission works and, finally, on the reinforcement of distribution networks to improve detailed supplies. For the purpose of this analysis assumptions have to be made regarding the additional personnel and equipment that will be required and the future organization necessary to achieve UFW control. It may be that the existing task force can be continued and expanded as necessary to undertake the additional works. Clearly there are great advantages in retaining the operation directly under the general manager's control for at least a few years, until a degree of- improved control has been achieved. But the works necessary impinge on the activities of the separate managers responsible for maintenance pumping, distribution, and finance. Close and continuous cooperation with these divisions wilL require the exercise of a considerable degree of tact for success. A better alternative might be to merge the UFW task force with the utility's regular maintenance activities and strengthen the entire maintenance effort. The permanent establishment of additional staff may entail government approval of a revised "establishment." The final task is the preparation of a program and budget for such additional investigations, operations, and new works as are shown to be justified bv the anticipated improvements. One factor to bear in mind is that the control of UFW is not a "once-only" operation, but a continuing process. 3-26 It must not be thought that after a great effort for a few years no further effort will be required. If the process comes to an end, the conditions will revert completely within a year or so. It could be that some variations in effort over the years may be economically justified. For instance, if new works have been completed, while there is surplus water available there may be less justification for high expenditure on leak detection and repair, depending on the marginal cost of water. But even if some easing off may be accepted, the basic aims and staff must be retained ready for reexpansion when conditions change. Distribution System Maintenance and UFW The UFW task force cannot achieve its objective without the cooperation of the distribution system engineer and his staff. The general manager of the utility should, at the same time he is supporting the UFW task force, upgrade distribution system maintenance standards over the same period of five years. At the end of the period, the comparative efficiency of the two units should be very carefully studied and their activities merged. UFW activities for the second five-year program would then be carried out by the distribution system engineer under the assistant manager for operations and maintenance. 3.14 Strategy It is assumed that from this point onward, UFW would be the responsibility of the assistant manager (O&M). The manager now has adequate data and is in a position to reconsider the overall strategy and to decide on a future provisional program and the measures to be adopted to implement that program. It may well be that this cannot be a final decision, and that the result of operations in the early years could justify some cutting back or expansion of the program. This will be possible provided all operations are closely monitored and evaluated continuously. Review program and budget to cover all actions described below, including preliminary estimates for long-term actions. On the basis of cost/benefit analysis, the general manager has to reconsider overall strategy and obtain the necessary authority and financial approval for proceeding with such additional works as are economically justified. Policy decision and implementation is now required, including confirmation of organizational changes. Decisions may include establishing the position of UFW manager (or equivalent) and asigning him a team, remuneration and incentive paymen s, and the appointment of consultants required for special studies.0 5/ UFW Manager could, after the first year or following achievement of results as head of the UFW Task Force, report to the head of the Department responsible for Operation and Maintenance. 3-27 Annex 3A. Typical Investigation of Consumer's Meter Accuracy and Domestic Consumption This investigation should be based on a 1 percent random sample of the total domestic meters in each zone or district. The account register is used for selection, taking 1 percent from each zone/district. At random, groups of about six meters are selected (rather than invididual meters) in order to reduce field work. From the account numbers, the meter reading books are used to provide the meter serial number, manufacturer, recorded consumption for the two previous quarters, and most recent date and reading. The selected meters are then site tested by a test team made up of an engineer/technician, a meter reader, and a meter supervisor. The test equipment consists of a pressure gauge and a test meter with filter, two lengths of hose, and couplings. The meter is located, read, and removed after noting the extent to which any isolating valves are throttled. The pressure is recorded. The meter is coupled in series with the test meter and connected by hose to the supply pipe and, for discharge, into the drain. The pressure reading will serve to check the zone in which the consumer is connected. The average flow rate over the past two periods is calculated and the meter tested at about that rate, and at one quarter of that rate. If the supply is indirect to tank through a ballvalve, the meter should also be tested at 20 liters per hour, which is the probable average flow rate for a closing ballvalve. The accuracy should be calculated as the mean of the percentage at average flow rate and one quarter of the average flow rate or, if there is a storage tank, at the mean accuracy at average flow rate and at 20 liters per hour. The supervisor should obtain details of the number of people served by the meter and any changes that have occurred. The results are analyzed to show the mean accuracy for each meter and the average daily consumption, as calculated from the last two meter readings and corrected for the inaccuracy now determined. Per capita consumption can also be calculated. The recorded and corrected consumption rates are totalled, and the average inaccuracy and average consumption in (liters per capita per day) is calculated for each zone or district. Where practicable, the results may be separated to determine per capita consumption in different social groups. 3-28 Annex 3B. Method of Locating Leaks by Sounding The location of leaks by sounding demands considerable experience, whichn can be acquired by good training in which the various problems are carefully demcnstrated. It is, of course, far easier to detect leaks by direct Lcntact with a main valve or service connections than through the ground. Some basic points on sounding techniques are covered in the following paragraphs. An ordinary metal rod or a valve key will transmit the sound of a leak. but extraneous noises make this simple sounding difficult. An improve- ment has been developed by fitting double earpieces containing diaphragms whichn amplify the sound. The sound can be detected by a microphone and amplified electronically. When this technique is used, leak sounds can be indi :ted visually and heard at greater distances. Traffic noises are deadened by rubber seals around the earpieces facilitating the detection of sound patterns that indicate types of leakage. Unfortunately, extraneous sounds other than traffic may also be amplified. All this equipment is valuable, but good training is essential for its proper use and above all for care ir handling to avoid breakage. Batteries must be checked and well na ntained, and all contacts must be cleaned to avoid corrosion. Differences in the type of sound will indicate to an experienced soeratsr whether he is listening to a leak or to water being drawn off for -onsumptiLon If the leak is near by, the sound is normally of a higher i-^e eney, Listening on a valve is often useful, provided the pipe material is to, asbestos cement or plastic. Leaks are easier to detect in metal pipes in asbestos cement and most difficult of all in plastic and rubber PZ;es. Leaks are more easily detected in dry soils than when the pipes are bel>o the water table, If there are air pockets in the pipe, sound is not tran-sm4tted beyond them. But the sound of a slight leak from a gland on a ve_ze will drown out the sound of a larger leak at a distance. The sound hea-- from hydrants or service pipes is much more faint, unless the leak is ass>.sated with the particular branch. It Is extrermiely difficult to locate a leak by reference to the intensity of sound at different points. Usually final location is best achl eved by indirect sounding on the ground or by listening on a pin driven -l t'o -contact with the pipe. For indirect sounding on the ground, it is usually best to use renones with dual hollow tubes connected to ground plates or to pointed pf-cbes ontaining diaphragms. This equipment is also available with birooones and ampLifiers. The plates or probes need to be vertically above -e >i,e thus it is usually best to begin by using magnetic pipe locators to detPminne the pipe location accurately. The sounds heard by indirect sounding are of a very low pitch and not easy to differentiate from traffic noises. vn:y by changing the position of the probe/plate in relation to the leak may dif-e ences be detected. Sounds emitted from metal pipes tend to be concentrated along the line of the pipe whereas those emitted from plastic pipes are spread uniformly in all directions. False indications can be obtained through transmission by foundation walls in contact with the pipe. 3-29 Depending on labor conditions, simple sounding -- without any means of concentrating effort. by district metering or otherwise -- aimed at covering the whole area of supply twice a year could entail the employment of one three-man team for every 20,000 properties. 3-30 Annex 3C. Costs of Various Components of UFW Control Task Range of Costs* (U.S.$) Sound 1,000 properties (three-man team) 200 to 400 Read 100 consumer's meters 20 to 50 Read 100 district meters 160 to 350 Install pressure tapping and construct chamber 400 to 600 Install district meter and set up district 2,500 to 5,000 Isolate a district and record night flow 100 to 150 Install PRV and set up zone 2,500 to 5,000 Undertake service reservoir drop test 50 to 100 Locate one leak with correlator 30 to 60 Replace consumer's meter 20 to 40 Associated large-scale work on meter (testing, repair, etc.) 50 to 100 Remapping with electronic tracing survey (per square kilometer) 2,000 to 10,000 (Alternatively, per kilometer of network pipelines) 100 to 200 For equipment allow $500 for a valve box locator, $1,000 for a pipe locator, and $20,000 for a correlator * These are based on 1987 prices. 3-31 A_nex 3D. Economic Assessment Models* General Water distribution systems can be analyzed by examining main break and leak patterns, corrosion history, and environmental stresses. By monitoring the distribution system, managers can assess its stability over time. The availability of this information greatly eases the manager's decision to repair, replace or rehabilitate water mains. In addition to this information economic analysis can also assist in maintenance planning. This section presents several economic analysis approaches for the evaluation of the repair, replacement, and rehabilitation options. There are a limited number of maintenance actions that can be used to preserve the structural integrity of a water distribution system. Several economic models have been developed to assist management decisions, based upon particular system conditions. The three maintenance alternatives for which economic models have been developed include: (a) Main Break Repair or Replacement (b) Cleaning and Lining (c) Leak Detection and Repair These alternatives and their associated economic evaluations are discussed in the following subsection. Main Break Repair or Replacement Water Main Break Costs There are several direct and indirect costs which can result from water main breaks. These include: (a) Direct water utility repair and emergency crew costs. (b) Direct water utility treatment and pumping costs associated with water loss from breaks. (c) Direct water utility expenses related to water damage claims. (d) Indirect city costs for utility overhead and for emergency police and fire protection during major breaks. (e) Commercial and residential water damage costs not reimbursed by the water utility. * Reprinted from Water Main Evaluation for Rehabilitation/Replacement, by permission. CopyrightO1986, AWWA Research Foundation, 6666 West Quincy Avenue, Denver, Colorado, U.S.A., 80235. 3-32 (g) Costs imposed on other utilities. (h) Costs of traffic and public transport disruption. These costs vary, from specific and quantifiable expenditures for repair materials and labor to more generalized and less quantifiable traffic and service disruptions. The water utility should record labor and materials used in break repairs so that realistic estimates of direct break repair cost can be made. WhiLe police, fire and streets department personnel often respond to main breaks, these detailed labor cost records are not usually available. Water damage costs from main breaks are not easily determined. For this reason, damage claim costs are not included in repair cost estimates presented below. Indirect overhead costs included in the estimates are for services provided to the water utility from other local government agencies, paid leave, fringe benefits, equipment and vehicle depreciation, in-house support (paperwork), maintenance, power, gasoline, and supervisory functions. There is no specific information on the extent and severity of indirect costs associated with traffic or service disruptions and nonreimbursed business and residential water damage losses. It is likely that the latter indirect costs greatly exceed the direct repair, emergency response, and claim settlement costs. Break Repairs The actual costs for water main repairs depend upon the pipe diameter and the type of break. Circumferential breaks are simpler to repair than longitudinal breaks. While a circumferential break can be repaired using a clamp around the broken pipe, a longitudinal break necessitates replacement Of a main section. Break-Even Analysis An economic model was developed to determine when main replacement becomes more cost effective than the continuation of its repair. The economic analysis for repair or replacement consists of determining the number of repairs per mile per year that can be performed before this break-even point is reached. This analysis is based upon direct labor and material costs for both repair and replacement as well as indirect over head costs. The economic trade-off between repair and replacement of a water main compares the initial cost of main replacement with the projected savings of future repair costs. Theoretically, the break-even rate may be defined as the point at which the cost of main replacement equals the future cost savings of break repair. If the occurrence of breaks on a particular pipe section exceeds the break-even rate, it is economically advantageous to replace the main. Conversely, segments with less than the break-even rate are not candidates for current replacement. The break-even situation is presented in Equation 10: 3-33 R = C (10) where: R = present worth of expected break repair savings. C = present cost of replacement. Equation 11 defines the present worth of main repair costs. R B x U x P(j,i,n) (11) where: B = initial main break rate; breaks/mile/year. U = cost of main repair for a single break. j = growth r.ate (break rate increase per year). i = net discount rate after accounting for inflation. n = period of study (years). P(j,i,n) = present worth equivalency factor (16) The present worth equivalency factor indicates the present worth at i percent interest of $1 of incurred cost at the end of the first year, and of amounts increasing by j percent from year to year until n years are completed. It is calculated by Equation 12. 1 P(j,i,n) =L - (1+K)"I (12) k where: K = ( l+i ) --1 ( l+j ) The break-even situation occurs when the present worth of repair cost savings (R) equals the replacement costs (C) as presented in Equation 10. The break-even rate (B') can be determined by combining Equations 10, 11, and 12 and rearranging terms. C B' = U x P(j,i,n) (13) where: B' = Current break-even main break rate such that the present worth of future break repair cost equals current replacement costs. Equation 13 can be rearranged to introduce a factor, F, which is the ratio of replacement costs/mile to the cost of a single break repair. F B' = P(j,i,n) (14) where: F = C/U 3-34 Figure 3D-i presents a plot of Equation 14 for five main break growth rates: 1%, 3%, 5%, 8%, and 10%. It can be seen that the initial break-even rate depends upon the replacement costs, repair costs, and the projected growth rate. This break-even analysis can be used on individual pipe sections to assess the cost of repair or replacement. As an example, consider a 3,000-foot section of 6-inch main which has had five breaks over the past two years. A 7% discount rate and a twenty-five year study period is assumed. This situation can be evaluated as follows: 1. Calculate initial main break rate: 5B/2yr B = 3000ft/5,280ft/mi B = 4.3 B/mi/yr 2. Calculate replacement-repair cost ratio: F = Replacement Cost = $400,000/mi = 100 B/mi Repair Cost $4,000/B 3. Determine break growth rate for utility: i = 1%/year break rate increase. 4. Determine break-even break rate from the ordinate of Figure 25: B' = 9 B/mi/yr 5. Compare break rate from Step 1 to break even rate from Step 4: 4.3 < 9 Action: Continue to repair main, do not replace. In this example, replacement of the main can not be justified since the break-even rate (B') was not achieved. The cost of replacement is greater than the present value of the future repair costs for a 1% growth rate in breaks per year and a replacement-repair cost ratio of 100. A break frequency growth rate of 8% or greater (see Figure 3D-1) is required to justify replacement. It should be stressed that these results will change if one assumes a different discount rate and/or study period. Sensitivity Analysis The sensitivity analyses confirm that the break-even break rate depends upon the ratio of replacement-to-repair costs, the break growth rate, the net discount rate and the planning period. The cost ratio was found to have the greatest impact. The break-even break rate was found to: 1. Vary directly and to be highly sensitive to the ratio of replacement and repair costs. Doubling the costs ratio results in a doubling of the break-even number of breaks/mile. 3-35 go ro~~~~~~3 _rw*ft i t itag M . . ~~~~as wo,t moo.... 3, 20~~~i o0 . to @te e so .. ., R* et IW1 tcao eweewanpo o wwa FOgOrg-P-1. Break-even n BreroRt H* ur 3-1 ra-en tO ln Grewe Rate fo ieAnul /tat ae 3-36 2. Vary inversely with the break growth rate such that higher growth rates require fewer initial breaks/mile before replacement can be justified. 3. Vary directly with the net discount rate in the 4-9 percent range such that increased financing costs permit more breaks/mile before replacement is justified. 4. Vary inversely with the planning period such that longer study periods allow fewer breaks/mile to justify replacement. The longer period favors early investment. Role of Indirect Costs The repair verses replacement cost comparisons described above indicate that the direct and overhead costs for main repairs are relatively minor when compared to replacement costs. These low break repair costs result in surprisingly high break-even break rates, which may be unacceptable from a public convenience standpoint. The direct and overhead cost comparisons yield these results because they consider only the easily measurable utility costs. They do not include the indirect costs of water and transportation service interruptions nor the unmeasured damages to other utilities and the road base. The true costs of water main breaks include these detrimental impacts. Unfortunately these costs are difficult to measure with data currently available. One strategy to account for these intangible costs is to develop a risk assessment mode. Water mains in select areas are prioritized for replacement according to a risk assessment scoring system. Scores are first assigned to each main based upon the potential for such hazards as traffic disruptions and damage from potential flooding. Weights are then assigned to each hazard based upon an assessment of the relative magnitude of indirect costs each contributes. A hazard assessment score is then prepared by multiplying the scores and weights for each hazard and adding up all these products. Mains may then be ranked based upon this score along with other considerations. Another strategy which may be used is to quantify these indirect costs on a cost basis using available data. These two methods have been used by several utilities to handle these indirect costs. Cleaning and Lining One available option to reduce costs for repair and replacement of water mains is to implement a rehabilitation program. In an attempt to increase the service life of unlined mains, address water quality problems, and improve hydraulic capacity, unlined cast iron mains can be cleaned and cement lined. Mains meeting certain pipe characteristics and maintenance criteria may be candidates for a cleaning and lining program (see Section 8). An economic model has been developed to evaluate the costs and benefits of cleaning and lining as a technique for prolonging the useful life of older mains. This evaluation is based on estimated material and labor construction costs for both options. The benefits of improved water quality, improved hydraulic capacity and increased fire flow capability from a health, safety 3-37 and public relations standpoint have not been quantified. These benefits may override the cost/benefit analysis thereby making rehabilitation of certain mains beneficial. Economic Model Cleaning and lining is attractive because it is considerably less costly than main replacements. The lower unit cost enables the utility to rehabilitate more miles of main with its limited capital funds. The economic model presented here has been developed to provide a method of evaluating costs of deferring replacement by cleaning and lining against the incremental costs of allowing the mains to continue deteriorating until replacement becomes necessary. It is recognized that cleaning and lining a main does not eliminate the need to replace it eventually. Rather, cleaning and lining is a method of prolonging the main's useful life and deferring replacement. The cleaning and lining cost trade-off presented conceptually in Figure 3D-2, shows the cash flows for both the rehabilitation and replacement options. The clean and lining option involves the initial rehabilitation cost and the replacement cost at a future year equal to the sum of the initial remaining main life and the additional service interval secured by rehabilitation. Replacement, on the other hand, involves the cost of replacement at the end of the main's remaining life. Cleaning and lining does not enhance the structural condition of mains, but it can prolong the useful life if the main has not deteriorated beyond an acceptable level. Ignoring break repair costs, the relative economics of replacement and rehabilitation can be defined by comparing the present worth costs for the two options. The present worth of costs for cleaning and lining may be estimated by Equation 15, below: PW-Reh = Rl + R2 (1+j)t (1+1 ) L(15) where: PW-Reh = present worth of cleaning and lining costs. i = discount rate. j = replacement cost inflation rate. Rl = current cleaning and lining cost, dollars. R2 = replacement cost in initial year dollars. t = year in which deferred replacement occurs; the sum of remaining life (r) and additional service period (p). The present worth of costs for the unaltered replacement option is presented in Equation 16, below: PW-Rep = R2 (1+j)r (l+l)r (16) where: PW-Rep = present worth of undeferred future replacement. R2 = replacement cost in initial year dollars. r = estimated remaining years in original main. 3-38 RHAELSTATION i 0 ecolaste I Ye ars RehatiEtatfon # C4st I ProloIed Year REPLACEMENT Replace Cost Year Figure 3D-2. Rehabilitation and Replacement Option Cost Trade-off 3-39 The ration of Equation 16 to 15 can be used to compare the options, as shown in Equation 17. K = PW-Rep PW-Reh (17) where: K = Ratio of the present worth of costs associated with replacement (Equation 16) to the present worth of costs for current rehabilitation and later replacement (Equation 15). Cases where the cost ratio is less than 1.0 indicate that future replacement is more cost-effective than rehabilitation. Figure 3D-3 presents estimates of the ratio K for various combinations of remaining life (r) and of prolonged years (p) due to rehabilitation. This figure is based on a net discount rate of 7% per year, rehabilitation costs of $35/LF, replacement costs of $125/LF and small diameter mains (6" and 8"). For this specific example, rehabilitation of mains with an estimated 20-30 years remaining life will never become economically favorable. These mains should be replaced rather than cleaned and lined. However, while replacement of these mains is economically favorable, the utility may not want to wait 20-30 years to replace them if severe hydraulic and/or water quality conditions exist. In this situation, the utility may decide to rehabilitate these mains immediately, despite the lower economic benefits, due to health, safety and public relations considerations. For mains with 10 years of remaining life, the return on the rehabilitation investment is achieved in the twelfth year. Sensitivity Analysis Ignoring break repair costs, the replacement/rehabilitation cost ratio K depends upon several factors, namely: (a) Remaining life in main (b) Life extension from rehabilitation (c) Discount rate (d) Replacement unit cost (e) Rehabilitation unit cost. The sensitivity of the replacement rehabilitation cost ratio to the above factors are summarized in the following paragraph. The K ratio was found to: 1. Be highly sensitive to both the estimated remaining life of the mains and the extension of life secured by cleaning and lining. (See Figure 3D-2.) 2. Vary inversely with the discount rate such that rehabilitation becomes less economically attractive with higher discount rates. This is illustrated in Table 3D-1 for the case of a pipe with twenty years remaining life and the same rehabilitation and replacement costs. As the discount rate is reduced frcmn 7% to 3%, rehabilitation becomes more attractive, especially as the years prolonged by rehabilitation increases. At the current 3% net discount rate, the curves shown in Figure 3D-3 would shift upward to favor a greater use of rehabilitation 3- 5 YEAR REMAININ0 2 _ o 10 YEARS RWMAINING u1 1 IREHAIBI-gTATION < o R~~~~~~~~~~~REAK EVEN ,, - - ~~~~~~20 YFARS REMAININGo 0~~ YEARS PROLONGED I2Y REHABILITATION FMr:e 30-3.S 30s YEati S to Prolo ngs! F0 - _____ Note: Based on net discount rate of 7% rehabilitation at $35 per linear feet and replacement at $125 per linear feet. Ratio is very sensitive to all three factors. 3-41 Table 3D-1. Replacement/Rehabilitation Cost Ratio Discount Rate Sensitivity for 20 Years of Remaining Life Years Prolonged Discount Rate % By Rehabilitation 3 7 5 0.73 0.57 10 0.80 0.64 20 0.94 0.77 30 1.09 0.85 40 1.23 0.91 50 1.36 0.94 3. Vary inversely to the unit costs of rehabilitation and replacement. This analysis indicates that cleaning and lining is generally more cost effective and the cost difference between the two alternatives increases. It shouLd be noted that the replacement option does not assume any cost for continued break repair. It should be noted that these findings do not apply to larger diameter transmission mains. The relative costs of rehabilitation of these mains are much lower. In addition, ihe energy savings from improved hydraulic conditions may make cleaning and lining cost-effective. Further, rehabilitation may be a worthwhile strategy due to the benefits of improved hycdraulic capacity, improved water quality, improved fire flow capability, and improved public relations. These economically intangible benefits have not been included in this mode. They are, however, an important part of the decision process. For this reason, a water main rehabilitation decision system may include these factors as criteria in the selection process. In addition, it should be mentioned that cement lining will neither renew consumer service connections, nor reduce external corrosion. The utility must still decide whether either of these factors should affect the decision to clean and line any particular segment of main. Leak Detection and Repair Along with addressing the costs of water main breakage, each utility should examine the cost effectiveness of leakage detection. The relationship between water main leakage and water main breaks has not been extensively explored. It is hypothesized that areas with high rates of leakage will also be areas with many main breaks. It is further hypothesized that these leaks are one possible cause of main breaks. If this is correct, an aggressive leakage detection program should significantly increase utilities' abilities to reduce the rate of increase in main breaks. By taking action before leaking mains break, the life of existing mains will be prolonged. The flow chart presented in Figure 3D-4 traces the procedure for comparing the costs and benefits of implementing a leak detection and repair 3-42 SYSTEM LEAKAGE ROUTINE LEAK ANO BREAK AGGRESSrVE LEAK DETECTION REPAIR PROGRAM ANO REPAIR PROGRAM [REDUCED SYSTE-M LEAKAGE COSTS COSTS LC6T WA ET & ANDp BREW~S REUE U AER TOM LEAK(S ANDO BREMS REDUCED LEM AND BREM REAJwS ROUTVE LEAK AND BREK RPAIR r LEAK DET!C7UN A eOST L D COMPARISON Figure 3D-4. Conceptual Illustration of the Leak Detection and Repair ModLel 3-43 program. If a utility chooses not to implement a program, its costs will include the cost of the lost water and the leak and break repair costs. When a utility implements at detection program and repairs the leaks discovered, the amount of lost water in the system decreases. The reduction in leakage may result in a reduction in break occurrences. The cost of the leak detection and repair activity must be compared to the benefits of a reduction in lost water and the possible reduction in additional break and leak repairs to determine the value of such a program. Leakage Control Model A leakage control model was developed to compare the projected future costs of a leak detection and repair program with the cost of lost water. The model has five elements, as follows: (a) Leakage in Base Year (b) Leakage Growth, percent/year (c) Leakage Located and Corrected Routinely (d) Leak Detection and Repair Program (e) Cost in Base Year Each element is discussed below. Leakage - this is the rate of water loss in a given year from all different sources. Growth Rate - new leaks form continuously. The rate of formation varies by leakc type and area. Routine Leakage Located - water utility crews repair many leaks throughout the year as part of their routine work. These leak repairs may result from customer complaints or from leaks found by crews in the course of other maintenance activities. Leak Detection and Repair (LD&R) Program - A LD&R program includes both the leak detection and repair components. Costs - leakage and its control includes three sets of costs: (1) cost of lost water, (2) cost of leak detection, and (3) cost of leak repairs. Main break costs are omitted from the analysis presented here since a precise relationship between break frequencies and leakage levels is difficult to determine. The leakage control model simulates the projected leakage condition over a user specified number of years under both control-program and no- program conditions. The economic trade-off between having a leak detection program or not must compare the cost of inspection and repair of detected leaks with the projected cost of lost water from the same leaks left unrepaired. If the cost of leaking water saved exceeds the cost of detection and repair, it is economically advantageous to proceed with a leak detection program. The situation is presented in Equation 18. 3-44 PWB = F2 - Fl (18) where: PWB = present worth benefit of detection program. Fl = present worth of costs with a detection program. F2 = present worth of costs with no detection program. Equation 19 defines the costs of the leak detection program for any one year, ignoring break rate and cost effects. wh = L x CR + L2 x S x CW + M x CD + x CR(19) where: Ll = leaks located by detection. L2 = leaks remaining after inspection. L3 = leaks routinely located. CR = cost of repair/leak. cW = cost of water. CD = cost of inspection/mile. M = miles inspected. S = size of leak (gpd). Equation 20 defines the costs for no program. F2 = 3 x CR + L2 x S x Cw + L, x S1 x Cw (20) The benefit in any year associated with a detection and repair program can therefore be stated as shown in Equation 21. B = Ll x S x Cw - M x CD L X CR (21) Sensitivity Analysis The present worth costs of the leak detection program and no-program alternatives are dependent upon several factors, namely: (a) Cost of water (b) Cost of repair (c) Leakage growth rate (d) Routine leak location rate (e) Net discount rate (f) Survey cycle The conclusions of the analyses are as follows: 1. As the cost of water increases, the cost of both the leak detection program and no program alternatives increases. As the water costs increase, the net savings for leak detection also increase. 3-45 2. The leak detection program is economically justified in nearly all cases of leak growth and leak location rates. See Table 3D-2. 3. The two alternatives vary inversely with the net discount rate. Although the leak detection program is the more favorable alternative, the net savings of leak detection decreases with the net discount rate. 4. The survey cycle only has an impact on the leak detection program alternative. It was found in the case examined that as the cycle changes, so does the net present worth of the program. A review of the net present worth costs under a variety of survey cycles can be used to determine the cycle where the program cost is the lowest and the net savings the greatest. A two-year inspection cycle for this situation was found to give the greatest savings. Future Detection and Repair Programs Using the economic model, future leak detection programs can be developed. The model in combination with a given utility's maintenance and cost experience, can be applied to develop a cost-effective survey cycle for each utility's survey district. To illustrate how the model can be used, a copy of printed results is presented in Table 3D-3. This illustrative application of the model addresses a twenty-year planning period, using data drawn from the accelerated program of leak detection in North Philadelphia (43). Discounted costs are shown, with and without a continuing detection program, for the assumption stated. 3-46 Table 3D-2. Sensitivity of Present Worth Benefits of Leak Detection to Leak Growth Rates and Routine Leak Location Rate Percentage Percentage Leak Growth Routine Leak Alternative with Lower Present Worth Rate/Year Location Rate Program No Program 1.0 1.0 x 1.0 3.0 x 1.0 5.0 x 1.0 9.0 x 1.0 13.0 x 2.5 2.5 x 2.5 7.5 x 2.5 12.5 x 2.5 17.5 x 5.0 1.0 x 5.0 3.0 x 5.0 5.0 x 5.0 9.0 x 5.0 13.0 x 7.5 2.5 x 7.5 7.5 x 7.5 12.5 x 7.5 17.5 x 13.0 1.0 x 13.0 3.0 x 13.0 5.0 x 13.0 9.0 x 13.0 13.0 x Table 3D-3. Sample Worksheet of the Leak Detection and Repair Model for Leak Detection Program and No-Program .Options ---1-15 -- 5-- -- --- I -- -- -- -- -- --- -- -- -- -- -- -- I CCstalktte $Igles Wase £1iU555C4 I. III MiCNtAIUt4 u 4e d 5.6.. uSkesi m mlks 0A .5 k as ll hi I h..6.. sit 55l. dim14 i two s.a BesIee 115 Av e $A se; NWat s Oak I~ 1#ils Su I SMal kte.5ta a.,,s, Cefi, real itlkile.s in I*Ae (p6.el Noto.. Nei e, * lasts SiSa seals tealsk h..sd eted S..td W4r Seals IsAlage tilI Wae# Lusts B Se l' s.,tdeatols leaSs" e 11 c tkshg call 51,5 15 534 55 9 41 4.5 4,2 l.ss 5211551565 Su I' AM6 sit 01.3 U 58 I 41 44 lU itS .0 Itt. ISaSl I' 59 9I AlUt J duS 61.1 ols.411 Bs.tni 941a.sz aIs& ini $-I It a4 it INs a.. as s SisIisAs's. 041.5, I Go 11 s af ss.e illiasee 0.1.U 5040,55 out4 Us .9 Is 55 1 Ut 0.611 to1 8.95 lit.w141 51)3, 114S.59 du(i SI is ct al 15. 51. 145ili O.&w, 5*41.i1 51,9 I'S 5.1 it Si if1 11 .5 1 1 .15 5s .f 114I1119.45 5366.511 I Ali SI to Al l i l 55.461,.16 l 0 , 1 61 4.161a 5396 l ~ .9 to isit 5 1. i) 11 41 4 *. tS 81 tlt 68I.41I 1516%I & SIs l, 0- it 1.1ii,il5 U -15 8. C.0 90 liutti ti to 85 3.5 I St if 6 1 .13 lie 1.31 1ea.396 S's.199 1t'e.ss i ll t a1 53 1 1 5.5 w. Si Isi .111 Iste.211 an, ins a..' 6 di s o' s #J ti as op 'no 1.Us 18M.isi 19.16 I nii is n il us., 019.41 114Bsse Owan$ Catsi tie 1.4 4 it it so 1.t9 CIts 9.516I Sloot C. la 110,141,I Of Al U VS 1ei ta- S ~ .ae5 616emS Bis,.i fill I?$ 3-4 4 5 4 I ItIAl535 2.51 fitW? 545,016 01.1.11 B1814I v Ad 4f 50 i i -Ii Csss.1i tes 11G.44N.lel tilt IllI t1. II Al Al to tie In .1,t Bess',41M6 l ips "A ist isCAl on5 i ti.s asian, Ct.ju susan 511 .113 a.a A 30 U do I Cis Illt 1.49 406.141 OWloS' 51.11511 i 01 aS 5 to At-s I el W3tISl 16040.111 ins Ill a. I Is if it 5.85 ftu u.54 514t.ils 516.45 1Isee' S "le is 4$ Sao Ic.a Buas,tte Stt.14a B.s:.SA til inl 4 I pi .U it i.it 5l3 2.65 06.6.16 111.861 0119.h4 I 939 if is gmi Cl. 14 .111 fit. at 16.4545 nos ti a.. of so is las1 fit ass1 fsll14 111,5 sus9 11d.14 5 to IS uel tOt1 141.011 111.411 *414,t15 ml, Is .9 1 1 o it S .31 CIa 1.99 151)11 Knoii Im.#)) 05,m It to ss Its B.as).4s W.uSI 544.41 ma isa ~i's C i1 is 5.05 011 lee (A will 601.j" lism". I o tel W St let ti-S W$.eSe 111.4tl 1,51.55 asi lie 5.1- to. s of 1.41 15%18.81 is~e98 sses.u'. 9:9w.6 a seots so sa lte) . III spa02 Baseac4 1111.11 al Iii 8.1 I Ad 51 to1 . B Ot ts its1 0etiw steities *aai,ne leD? 64 It aae It.) 014.ts1 0.166ai Bieo.8 3-48 Annex 3E. Economic Analysis of Leak Detection* General Leakage represents a physical loss of water, but the contro; of leakage entails some cost and normally can be justified only in relation to the cost of producing the water saved by leakage reduction. Additional savings may be achieved if leakage control results in lower costs of sewerage and sewage treatment. Where water resources are limited, the simple anaLysis of cost of water saved in relation to the cost of leakage control is not the sole justification for active control. Furthermore, the active control of leakage entails improved control of the water system and is of direct benefit to management and thus to the cost of supplying water. Leaving aside these associated aspects not readily assessable in economic terms, it is relatively simple to undertake a cost/benefit analysis of leakage control and thus to determine future policy regarding the degree of control to be exercised. The cost of leakage is reflected in (a) the variable element of operating costs and (b) the capital costs of new works, to the extent that the construction of new works may be deferred. When an authority has a large surplus production capacity and major distribution works, and does not foresee incurring large capital costs in the near future, increased expenditure on leakage control may be less justified. Nonetheless it must be remembered that a reduction in leakage can help to increase revenue by increasing water sales. Cost of Leakage Unit cost of leakage = unit operating cost + unit capital cost Unit Operating Cost To calculate the unit operating cost (marginal cost) of water, one is concerned only with those water sources where production costs would be reduced by reducing leakage. Normally the large gravity source or the locally pumped source will produce the least costly supply and will be retained. Bulk supply sources or more recent and more remote pumped sources may entail more expensive pumping and treatment costs where reduction in output will be of greatest benefit. When reductions can be foreseen at more than one source it is necessary to calculate the weighted average of the variable operating costs at each source. To calculate operating costs, fixed costs that are not affected by variations in quantity produced must be included. The normally calculated total operating costs of water from each source station are not relevant. Variable operating costs include electricity for pumping and chemicals for treatment. Other variable costs usually are not significant. Reproduced with permission of the Water Research Centre, Henley Road, Melmenham, P.O. Box 16, Marlow, Bucks. SL72HD, England. 3-49 Operating Costs To calculate electricity costs if there is no power meter, use the following formulae: power input = 1.73 x volts x amps x power factor 1,000 and unit pumping cost = power input x kwh unit charge volume pumped (m3/h) unit treatment cost = annual chemical costs annual quantity of treated water NOTE The operating costs are reduced to currency unit/m3 Inflation factors should be ignored because they will apply more or less equally to labor and other costs of leakage control. Capital Costs Normally projections are made for five-year periods, but they can be made for a more extended period if large expenditures such as new source works or transmission mains are likely. It is necessary to estimate for each year the capital expenditure for source works, treatment works, pump stations, service reservoirs, transmission mains, primary distribution mains, and distribution reinforcements. The capitalized fixed costs of operating the future new works must be added to the capital costs. These costs are estimated by reference to the fixed costs of existing similar works; annual costs are capitalized by applying the factor (1 + r)/r, where r is the discount rate. For example, for a 5 percent discount rate the factor is 21. The future capital costs are entered for the appropriate year as shown in an example in Figure 3E-l. The diagonal lines in the table provide for insertion of the simple capital sum and a sum modified by applying multipliers (in the range 0 to 1) as indicated below. Where a project cannot be deferred for some good reason, it may be wholly or partially omitted and in this analysis a "demand multiplier" should be applied. Where a new project will satisfy demand for a long period ahead, its cost should be increased using a capacity multiplier, derived from Figure 3E-2. Using Figure 3E-2 the total discounted capital cost (TDCC) can be obtained, and the unit capital cost is then calculated from the formula: Unit capital cost = TDCC x r2 (l+r) x 3.65 x d r = the discount rate d = the projected annual rate of increase in demand FIGURE 3E-1 TYPICAL EXAMPLE FOR CALCULATION OF TOTAL DISCOUNTED CAPITAL COST (TDCC) Actual - ~ Costs of Capital Works (x £ 1000) u ,Modified Year modified Discounted Yea r Storage and Pumping Treatment Capital Discount Capital Distrtbution Station Works Mains Source Trunk Cost Factor Cost Extensions Mods Extension Strengthening Works Main (x £ 1000) (5%) (x £ 1000) i il iii iv v vI vii viii ix x xi Xli xiii six ° XV : 178 , 82 6 - -- -s 500 100 500 " 9338 7156 76 50 7 7 0 7 / 74 - / 80 1 7 6 / 7 204 095 194 141 .- 63 7 - 2 106 79 96 /z 8 106 /7 79 , 96 414 09 377 81 7 z 7 201 7 19 85 128 3- 7 ~~~~~6 98 82 136 0.86 117 6 130 4eulmtipinietr 1 178 83lp I-l. 1 9 1 191 0.82 157 83'191 12 500 370 91 1871 0,56 1048 1075 796 13 5x2 I 92 226 0.53 120 226 Demand multple 10 1.0 05 07 1 0 '1.0Toa £25100 Capacity 1 9251 1 1 Multiplier 191921 521 World Hank--30891 I 3-51 FIGURE 3e-2 Capacity Multipliers for Various Discount Rates 4.5 85% 4.0 3.5 7% 3.0 0 o 2.5 0 U 10 1.G 1.0 5 10 15 20 25 30 Years of Demand Increases Satisfied by Works World Bank-30505 3 3-52 Cost of Leakage Control Having determined the extent of leakage and the unit cost of leakage, management has to set a realistic target for leakage control. This may be a reduction of UFW from say 40 percent to 30 percent (or it may be expressed in Liters per property hour as given by net night production). The cost of improved leakage control based on increased labor and equipment then has to be estimated based on the type of control proposed. The results must be monitored to ensure that the target is being achieved without the cost of control being exceeded. The step-by-step procedure proposed in this document ensures that improved control is maintained within the economic limit justified by the saving in cost of water produced. Costs of leakage control can vary greatly depending on: labor costs in the particular country; extra costs for night work; and above all, the condition of the system at the outset. If the system has become very run-down due to negligible input on proper operation and maintenance, the preliminary cost of putting the system into reasonable condition could form a considerable part of the total. Assuming that such preliminary work is charged to general operation and not to the waste detection program, typical costs of various components would be in line with those noted in Annex 3C. The costs of associated repairs should be readily determined from existing records but, failing that, might be based on: a team of five, plus a driver; repairs to pipe (five hours), to service (three hours), to valve or hydrant (two hours). General Comment The initiation of active control can be expected to show a considerable financial return but, as the program is developed and leaks are repaired, the results gradually become less rewarding. The law of diminishing returns applies, and it is advisable to repeat the economic analysis at regular intervals using updated cost projections to ensure that increasing efforts are fully justified. It must not be overlooked that the most important part of control is to repair rapidly the leaks when they are located. For a detailed discussion the reader is referred to Standing Technical Committee Report No. 26 on Leakage Control PoLicy and Practice published by the U.K. National Water Council. 3-53 Annex 3F. Draft Typical Terms of Reference for Consultants Background (Brief description of the Authority's organization and supply system.) Objectives The Authority proposes to undertake a study of the whole distribution system from source stations to consumers' fittings with a view to reducing the unaccounted-for water (UFW) and improving control of the system. The UFW appears to be excessive partly due to defects in consumers' meters and certain illegal practices and partly because of losses by waste including underground leakage. The aim is to reduce metering losses by about not less than 15 percent within a period of three years and losses caused by all forms of waste by about 25 percent within a period of five years. The Authority proposes to appoint consultants to assist in various aspects of the study in order to implement the procedure required to undertake the improvements necessary to enable the above objectives to be achieved and leave in place a sustainable organization. Consultants will also be hired to make such additional reductions in UFW in future years as may be economically justified. The attached "Working Guidelines" are provided for your review and comments. The Guidelines would then form part of the terms of reference. Consultants will be required to assist the Authority in the following aspects of the work: 1. The setting to work and calibration of production meters and large consumption meters, including the correct installation of new meters where required and the regular monitoring of all water delivered by these meters. 2. The reassessment of total UFW, 3. A survey of all illegal connections, followed by the removal or the metering of such connections. 4. A sample survey of the accuracy of consumers' meters and of any irregularities in meter reading in order to assess (a) the total loss attributable to metering deficiencies and (b) the actual unit domestic consumption in different social conditions and supply areas. 5. The allocation of total UFW between metering losses and physical losses due to waste. 6. The training of selected members of the Authority's staff in the practical and administrative aspects of reducing waste. 3-54 7. A review of all aspects of metering practice, including possible privatization, incentives to improve performance and, if required, design and supervision of extensions to the meter repair shop. 8. A review of present waste control practices and an econonmic study of alternative methods for improved waste control, followed by advice on future policy. 9. The survey and remapping of the complete network of mains and consumer services and the establishment of a system to ensure continuous updating. 10. The physical separation of the network into isolated districts serving about 10,000-to-20,000 properties. 11. The measurement, monitoring, and evaluation of rates of flow into the whole system and into each isolated district to facilitate the determination and subsequent control of total waste in each district. 12. The establishment and operation of an active waste detection and repair organization, including the tracing of leaks and all forms of waste by appropriate methods. 13. The institution and enforcement of bylaws for the control of waste water by establishment of a plumbing inspection division or other appropriate measures. 14. A network analysis of the distribution network to determine the status of amount operating conditions. 15. A preliminary economic study of the least-cost network reinforcement required to meet demand foreseen in existing planning proposals for the next ten years. 16. Continued advice, training and management services for a period of (three?) years. Submissions Consultants with experience in the relevant administrative, financial, operational, and maintenance problems are invited to submit proposals for undertaking all (or any) of the above, including provision of the necessary services, equipment, and labor for: Phase I Works covering a period of up to one year, and Phase II Works covering a further period of four years. Consultants should consider collaboration with local consulting companies in order to achieve greater understanding of local conditions and also to disseminate knowledge and experience more widely to achieve more permanent improvements in the maintenance of the supply. The proposals for labor, equipment, and time required (a) by the consultants and (b) by the Authority are to be itemized for each of the works described in the following sections. 3-55 Scope of Consultants Services Phase I Works Large Meters. In some cases, production meters exist for measuring flows of treated water from certain works but are not operating satisfactorily and large consumer meters may be under recording or be completely inoperative causing significant revenue losses. All meters have to be inspected and, where practicable, put into working order and calibrated at all relevant flows from the station. The calibration should, preferably, be undertaken by measurement of volume changes on suction or delivery tanks, or alternatively, by discharge through test meter or such other means as may be agreed upon by the Authority. Where meters cannot reasonably be put into working order or where meters do not already exist, new meters are to be purchased and installed at the Authority's expense in such a way as to minimize interference with the supply of water. Sufficient meters are to be purchased and installed to enable the total rate of flow from all source works into the distribution system and use by large consumers to be measured at all times. Reassessment of UFW. Having determined the true figure for total quantity of water produced and delivered into the system, it remains to calculate the total quantity that can be accounted for in supplies to the consumers. Two factors must be kept in mind: (a) the quantity supplied to consumers or used for public purposes but not metered and (b) the lack of correlation between the dates of reading various consumers' meters. A study is required of the seasonal or other variations in accounted for water compared with the water produced. The total UFW for the supply area can then to be calculated as a percentage of the water produced. A survey is also required of the number of properties in each district so that UFW can also be calculated in terms of quantity per property per hour. Comparisons should be made of these quantities in the different districts. Illegal Connections. (Normally applicable only when domestic supplies are metered.) A sample survey must be undertaken of the total number of properties receiving water supply compared with the number of metered connections. This survey should cover different social areas separately to determine the extent of illegal domestic supplies and variations between different areas. It must cover supplies afforded by neighbors through temporary house connections as well as permanent connections. The survey must show the number of inhabitants (as well as properties) receiving an illegal supply. In addition, the survey must cover industrial premises, which may have some unmetered, as well as metered, connections. Where consultants propose to use aerial sewage students or local labor for such surveys, their proposals must be detailed for independent checking to minimize fraud or other inaccuracies in data obtained. Consultants are also required to advise on appropriate steps (bylaws, tariffs, etc.) to be taken to minimize the reconnection of illegal supplies that have been cut off and to ensure that all connections are and remain metered. Sample Meter Accuracy and Domestic Consumption. A study of the accuracy of existing meters should be made, based on testing a 1 percent random sample of installed meters by means of a portable test meter. The site data must include data on the number of inhabitants served from each meter and the setting of the consumers' control valve when first examined. The accuracy 3-56 tests must especially include calibration at the flow rate for this normal valve setting, as well as for very poor flow rates. The Authority's meter readers and other assistants will be made available as necessary for locating meters and other assistance. The accuracy of meters should be recorded for each manufacturer and type of meter as well as (where the Authorityts records are available) the time that has elapsed since the meter was (a) last read and (b) last changed. In addition, information from the last two meter readings for all the meters tested is required to determine total metered consumption through the meters tested, total consumption per capita per day, and, where possible, comparable figures for different districts and different social groups. This study must also investigate any possible irregularities in meter reading, estimate the total effect of any such irregularities, and advise on measures to be adopted to reduce the problem. It should also be aimed at determining the extent of water wastage, including underground leakage, on the consumer's premises. Nonphysical UFW. Consultants will be required to determine from their surveys of meter accuracy, illegalities, and irregularities the actual quantities of water being delivered to consumers and thus, how much of UFW is attributable to metering deficiency. By deduction from total UFW, that part of UFW attributable to leakage should be calculated. Training and Equipment. Consultants are required to provide specialized training of not less than 20 weeks for (six?) of the Authority's employees in all aspects of waste detection, leakage repair, and prevention of waste, including inspection of consumer's plumbing and fittings. This training should be undertaken by the consultants' staff in association with employees of a water authority in a developed country. It should include lectures on theoretical aspects but should be particularly devoted to practical training under working conditions using traditional methods as well as more modern equipment for leak detection. This training abroad should be followed by further training by the consultant on the Authority's works. It initially should include assistance on the works described in Sections 4.1 - 4.5 and 4.8 of the Working Guidelines and, later, on the proposed Phase II Works. Consultants must provide the specialized leak detection equipment required initially and the specifications for all additional equipment required for implementation of the proposed policy. As part of the extended training, consultants are required to establish a pilot waste meter district, which will be regularly monitored and studied by all trainees. Metering Practice. Following the samDle tests on consumer meter accuracy, a review is required of present metering practice. This review should focus on the type, diameter, and manufacture of meters in current use; the required level of meter testing, repair, and replacement for effective operation; the possible alternative use of flow limiters for domestic consumers to ensure continued water conservation; and means to simplify maintenance, operation, and billing procedure. The quality of the water supplied, particularly the suspended solids, and the potential for improvements in quality by better maintenance of the network (including increased flushing, rehabilitation by cleaning, and relining of pipework) are all factors to be taken into account. 3-57 The practice for replacement, repair, and testing of meters should be examined and recommendations should be made as to possible incentives for improved output and changes in frequency of reading, replacement, and associated extensions of repair and testing shop facilities. A special study should be made on large meter installation, calibration and maintenance practices. Attention is to be given to possible privatization of meter reading and billing, which may also include privatization of meter installation, maintenance, testing, and repairs. Depending on the conclusions and recommendations reached in regard to metering practice, the consultant may be required to design and supervise the construction of extensions to, and management of, meter repair and testing shop facilities. Waste Detection and Repair Policy. In collaboration with the Authority, consultants are required to (a) review present waste control practices, (b) consider alternative, systematic methods required to meet target reductions in physical UFW, (c) examine the costs of waste detection methods, (d) undertake an economic study of alternative methods, and (e) advise on the policy to be adopted and the administrative and organizational changes to be made for the policy to be implemented, including the establishment of a practicable bonus scheme to stimulate enthusiasm among the inspectors and other personnel engaged in the works. Phase II Works Network Survey and Remapping. Existing maps of the distribution network are probably no longer accurate; many valves, hydrants and fittings have been covered over and can no longer be found; many are not operative. Specialized services are required, using electronic methods, to trace all mains and service connections and to plot on new maps to a scale of 1:1,000 showing correct sizes, materials, positions, and depths of all pipelines at 33 meter intervals. The maps are also required to show topographical details with ground elevations to 0.1 meter accuracy at all major pipework connections and otherwise at about 100 meter intervals. This work is to be undertaken in sections beginning with the older parts of the network and eventually covering the whole network. An organization should be established to ensure that all future extensions, reinforcements, and modifications are recorded as constructed and that all relevant data is transferred to the new maps. The adoption of a computer mapping system should be considered and approved by the Authority. Pressure Zones and Districts. The existing boundaries of pressure zones and districts should be examined with a view to reducing any excessive pressure that may exist and increasing pressure where supplies are inadequate. This process should also ensure that the rates of inflow to each district can be monitored at all times. Where districts have not been established, they should be set up; normally they should not contain more than about 20,000 properties. First, the proposed district boundaries should be 3-58 determined. When they have been approved by the Authority, a project for installing the necessary valving, metering, and modifications should be designed and specified; construction should be supervised and, finally, tested to demonstrate integrity. The district is to be surveyed demographically, and a monitoring system is to be established and maintained in collaboration with the Authority to enable changes in unit and total consumption to be regularly evaluated. Waste Detection Division. Consultants are required to advise on the setting up (or the re-constitution and enlargement) of active waste detection and repair services. Appropriate organization and staffing, designed to implement the proposed policy as determined by the economic study undertaken in Section 4.8. of the Working Guidelines should be included. Consultants are also required to provide management services to assist in the operation of this division for a period of about four years and to ensure its continued operation thereafter, as may be dictated by economic policy, so as to increase the control of all forms of waste water. The consultant's duties should include, but not be limited to, a review of, and advice on, current procedures for applications for new water services and on customers' internal service and plumbing repairs. Bylaws. The Authority proposes to request the Government to incroduce legislation for the introduction of (or alternatively to reactivate the operation of) bylaws for the control of waste water. Consultants are required to advise on all aspects of the implementation of this policy, including the provision of a plumbing inspectorate to ensure compliance with required standards of fittings and plumbing and with the specifications for all new main laying and repair works. Particular emphasis has to be placed on public relations and, if necessary, on proposed enforcement methods that may be necessary to ensure compliance with the bylaws. Network Analysis. In order to improve operation and control of the overall program; that is, to extend control of UFW and to determine least-cost network reinforcements to meet planned expansion of the supply, the Authority proposes a computer analysis of the existing network. Following the network survey in Section 4.9 of the Working Guidelines, consultants are required to undertake the necessary network analysis and to ensure correct calibration that represents current conditions. It is expected that certain anomalies may be discovered and that modifications of the network may be advisable in order to simplify operation or reduce the range of operating pressures and thus to reduce waste. Consultants are required to examine all possible modifications and, subject to an economic study, to advise on and to design and specify works for metering those changes that are justified. The network analysis is amply justified to achieve effective control of UFW as well as improved knowledge and operation of the system. Network Reinforcement. A natural corollary not strictly related to UFW control is the use of the network analysis for least-cost reinforcement of the network in order to meet planned expansion of water supply. This work may be conveniently included in the terms of reference for consultants, who also may be called upon to design and supervise the construction of new main laying work and to ensure improved standards, thus preventing perpetuation of excessive underground leakage. 3-59 Continued Advice on Maintenance and Associated Management Services. The Authority is aware that the problems of excessive UFW are only one example of inadequate attention to maintenance of all parts of the system resulting mainly from lack of sufficient trained and experienced personnel. The aim is to correct this situation by improving the training of personnel at all levels and thus improving the management and organization so that a higher level of service and a more reliable supply of water is made available to meet future needs. Consultants are required to maintain a presence of experienced operators working in the Authority for a period of about three years after the conclusion of the contract covering the above works. These operators are to be engaged in in-house training of personnel, advice to the Authority on improvements of operation and maintenance, and ultimately in achieving independence of consultant assistance. Reports Consultants should provide the Authority with monthly progress reports during Phase I works, together with a separate report at the conclusion of each of the itemized works. Consultants are required to prepare, in consultation with the Authority, monitoring indicators and a reporting system designed to provide clear information on the physical achievements and overall progress completed during Phase II works. At the conclusion of these works a comprehensive report is to be furnished giving full details of all data collected and all accomplishments. The reports should have a succint executive summary (both in English and the local language) which should have clear conclusions and recommendations. All reports are to be presented in English (30 copies) and in the local language. The number of local language copies should be decided in consultation with the client. 4-1 SECTION 4 EXTENDED ACTION PROGRAM FOR REDUCING NONPHYSICAL LOSSES 4.1 General In the past, some authorities have had serious problems with meters owned by consumers, and this practice has hopefully now been superseded. The consumers' meters must therefore be owned by the authority, which will set an appropriate rental charge to cover the initial cost, installation, and maintenance costs. The choice of meter type depends on the quality of the water but, for all large industrial or small domestic purposes, rotary piston meters are normally preiferable to inferential turbine (multijet or single jet) meters because of their greater accuracy. The meters should in most cases, be protected by additional strainers to limit erosion or seizing of the moving parts. However, if the water supplied contains large quantities of sediment, sand, or rust it will seriously interfere with rotary piston meter performance and life. There could then be a justification for using inferential turbine meters. 4.2 Domestic Meters Domestic meters are normally 15 millimeters in diameter, but some authorities have standardized on 22 millimeter diameters because of past low pressure problems. Rotary piston meters are normally constructed to comply with I.S.O. 4064/1; inferential (turbine) meters should comply with American Water Works Association Standard C-708. (European Economic Committee and other acceptable standards also exist.) Meters are subject to gross inaccuracy at low flow rates and this inaccuracy increases with age long before the meter fails completely (see Section 3.10). Where past supplies have been discontinuous, the practice of installing very large storage tanks to cover periods of deficient supply has evolved. This has encouraged the practice of throttling the supply so that the tanks are filling slowly over most of the 24-hour period, with the result that underregistration prevails at all times. A special charge could therefore be considered for consumers with individual storage tanks. If at all possible, the authority should aim to purchase all large tanks and provide smaller replacements once a continuous supply under pressure can be assured. Alternatively, the tank could become redundant, or its capacity could be reduced with volume displacers; such as concrete blocks or bricks. Recent research has shown that for all normal requirements 15 millimeters is unnecessarily large. It provides for normal flow rates of 1,500 liters per hour (the 22 millimeter diameter will pass up to 5,000 liters per hour) whereas 750 liters per hour is adequate where there is indirect plumbing and could be obtained with i0-;2 mnijlimeter diameter meters. The same findings were reported from Brazil, where smaller prototype meters with 15 millimeter diameter outlets have been produced. Tests show that, whereas with 15 millimeter diameter meters the average underregistration of new meters in normal use exceeds 5 percent, with smaller diameter meters of 750 liters per hour rating, the average underregistration is about 2 percent. 4-2 The "small body" domestic meter with a rated flow of 750 liters per hour may be used in the future to minimize underregistration at low flows, but initially it should be tested on a small scale. The problem is that the moving parts have smaller clearances and the meters may therefore be more prone to failure with water that contains excessive matter in suspension, which cannot be eliminated by protective screens. Such meters are in production at a slightly higher cost, which would be offset by increased water sales and, in any event, this extra cost might be reduced by increased orders. Selection and procurement of water meters is a complex matter and requires assistance from speciaLized consultants. When an authority reports that siv 25 percent of its meters are inoperative, it is probable that at least another 25 percent are recording only a fraction of the flow. When a meter is not operating, the practice is to charge based on an estimated flow. This estimate is based on registration recorded before the meter failed, and so is likely to be a very low estimate. Lack of proper maintenance of meters is thus a prime cause of nonphysical loss. 4.3 Large Meters In most cities, whether or not domestic supplies are metered, a very large proportion of the water consumed is sold through relatively few large diameter meters for bulk supplies to neighboring authorities or to industry, commerce, military camps and large government institutions or other public entities. It is of great importance to ensure that these large quantities are being recorded correctly; as they are few in number this is a relatively easy task. The first and most essential task is to site check whether these meters do record normal and low flows within the normally accepted tolerance of +2 percent. To ensure this, the meters must be tested frequently (at least annually) for accuracy. They must also be changed, cleaned, repaired, adjusted as necessary, and shop tested before reinstallation. These tests must be undertaken at both very low flows and normal flows. Although accuracy at low flows is difficult to ascertain, low flows may persist for long periods and represent large quantities of water. The larger the meter the more frequently it should be site tested and, if necessary, changed for shop refitting and testing. Depending on the quality of the water, routine replacement of the very large meters every two years may be found cost effective. The size of the meter should be appropriate to the maximum rate of demand without unacceptable loss of pressure. When demand is very high, compound metering may be necessary. Industrial demands fluctuate with business cycles; where a decline in business activity and in the rate of demand has occurred, a change to a smaller meter may be warranted. Otherwise, an unnecessarily high proportion of the flow may be subject to excessive underregistration. An independent historical check should, if possible, be carried out to ascertain how consumption relates to the volume of business. If at any time the increased activity is not reflected in increased consumption, an explanation should be sought. Likewise, an increase in metered demand without a corresponding increase in business activity should be investigated. It could be due to a major leakage or other type of waste on the consumer's premises. Irrespective of changes in demand, an independent check is justified to confirm that the size of the present meter corresponds to the maximum and minimum rates of flow recorded. The site testing of these meters may be by isolation and timed volume change in the consumer's private storaze. If this is not possible, provision should be made when the meter is instailed for a spot test at any time by discharge to waste through a test meter. Valves closed for this purpose must be checked to ensure that they shut off flow effectively. The choice of manufacturer and type of meter is important. For most purposes displacement meters are favoured for greater accuracy but, where failure of supply could have serious repercussions (for example, at a hospital or where expensive material is being produced), local storage is essential. If not possible, an inferential turbine meter would be required, which probably would be less accurate at low flows. 4.4 Meter Reading Procedure While the maintenance of meters, including changing, repairing, and testing, normally falls within the duties of the distribution manager, the reading of the meters, together with billing and collection of charges, is the job of the financial manager. A common procedure is for the authority's area to be divided into meter districts; each meter reader covers a block of properties within that district. Sometimes there is provision for interchange of blocks between the different readers. Quite often meters are read and accounts sent out at bimonthly or quarterly intervals. The actual reading dates are spread over the whole quarter, but the period between readings is about 90 days. For this reason, a comparison between production and consumption figures over a short period is not valid unless special correlated readings are made for that purpose. Although the reading of domesti:c meters is reasonable at bimonthly or quarterly intervals, this may not be the case for the larger industrial meters. Investigations will show whether it is not more cost effective to read the larger meters every month. This not only would permit improved collection, but would also ensure that any defective large meters are replaced with minimum delay. The larger the meter, the more frequently it should be read. Meter failures frequently result from excessive grit and rust particles in the water. This usually occurs because mains are not being flushed systematically. Flushing is most effective from washouts discharging to drains, ditches, or culverts, but flushing from hydrants is necessary where washouts do not exist (Hydrants are essential on all dead ends of mains and should be spaced about every 100 metres.) Irregularities in meter reading procedures can be rapidly detected if computerized accounting is introduced and programmed to throw up such anomalies. Failing that, frequent spot checks by management may be necessary. Concentrate on large meters and institute program of (a) more frequent reading according to size; (b) regular (annual) testing on- 4-4 site or replacement; and (c) critical examination of records and follow-up. Consider monthly reading and quarterly testing of meters 3 inches in diameter and over. Study general meter reading procedure and results. Undertake spot tests by management for sample study of reliability of meters and meter readings. Undertake sample survey of illegal connections in selected areas. Take steps to eliminate irregularities and anomalies. Establish a program for regular (four-to-five year) meter changing and testing. Commence color coding or similar identification method. For this purpose a sample study of the meter condition is required; it also detects any installation problems. Examine records of meter failures and failure to register low flows. Consider standardization appropriate to the operating conditions of type, and normal domestic diameter to minimize low flow errors. Examine a mains flushing policy and water treatment procedures to improve water quality and reduce meter failures. Alternatively, consider the use of "flow limiters" instead of meters. A pilot study should be made because some flow limiters are useless under conditions of low pressure. Determine policy and purchase additional or replacement meters necessary to meet future need. A local university might be approved to undertake these tasks as a research project. NOTE: The reading, billing, and collection procedures, including costs, should at this time be reviewed by management, with technical assistance as necessary, and designed to improve data retrieval, as well as to highlight leakages on consumers' premises, and reduce irregularities, underregistration, and illegalities. 4.5 Meter Repairs and Testing The procedure of changing meters at regular intervals requires the existence of a meter repair shop for cleaning, repair, and testing of defective meters. To be effective, the annual output of reconditioned meters must be about 25 percent of the total number installed (and 50 percent or possibly 100 percent of the total number of large meters). If the current capacity is inadequate, the location of the site in relation to the center of demand, traffic probLems, and journey times should be reviewed before expansion is planned. Having established the best location, capacity may be determined on the basis of two-to-three square meters for every meter to be serviced per day. This space will allow for receiving old meters, cleaning, repairing, painting, storage, testing, and dispatch, as well as office space for records and storage for new replacement meters. It is essential to ensure that accurate testing is done at very low rates. For domestic meters and indirect tank supply plumbing arrangements, a low rate would be about 15-to-20 liters per hour; for larger meters, proportionate increases in minimum test flow are allowable. The recorded flow 4-5 should not show more than +2 percent difference from the flow as measured by tank volume changes. One test line may be expected to cover about 50 domestic meters a day. Many of the meters will require only cleaning but experience indicates that between 10 percent and 20 nercent will probably require repairs. It is recommended that aLl meters be painted a distinctive color to denote the year they wiere reconditioned. This will help to ensure that all meters are changed at regular intervals. A study of the records Df meter breakdowns and repairs is a necessary preliminary to the formulation or policy on the selection, installation, and maintenance of meters. If no records are available, they must be started as soon as practicable. Records should be kept on cards or in a form by which data can be rapidly retrieved to show manufacturer, type, size, frequency, and nature of failures that have occurred. If records show very frequent failures because of suspended solids in the water, the use of modern flow limiters through which the flow rate remains fairly constant over a wide pressure range should be considered, and a pilot study should be made. A study of output per person may be necessary in order to ensure that an expanded meter workshop is designed to minimize the cost of repairs. This study should include the cost of meter replacement, including transport to the workshop. Study meter repair and testing shop procedure and capacity. Amend procedure as necessary to ensure testing at very low flows. Consider incentives for improved output of testing and repair. Study output per person aind future manpower requirements. Study records of meter failures and establish necessary data retrieval procedure for reducing failures. Design any necessary expansion of meter testing and repair workshop. 4.6 Boundaries of Districts The disparity between boundaries of meter districts and distribution districts often causes problems. Because of this difference, no accurate comparison of district production and demand can be made. This problem may be overcome by a one-off survey that allows the distribution district number to be entered on each meter account so that, with computerized billings, meter consumption for each distribution district can be readily totalled. Where billing is not computerized, correlation of boundaries becomes a much more difficult task that could entail physical changes in the boundaries of meter districts and even of meter readers "blocks." Moreover, if changes in distribution zones are required in order to reduce leakage by recducing pressure on consumers' services, a change in the meter reading district boundary may become necessary. Such changes may not be easy to achieve. 4-6 Examine all possible means of correlating meter consumption statistics with distribution district boundaries, including any necessary boundary changes. 4.7 Review and Further Action Program After consideration of the various factors involved in providing and reading consumers' meters, as well as any associated needs for additional or replacement meters and facilities for the servicing of meters, it is time to evaluate any improvements that have been effected and to consider any additional changes that appear necessary. Such changes might include: further modifications of the -eter reading procedure, upgrading the meter reader's responsibilities and rewards; arrangements for replacing meters; meter workshop practices and requirements for expansion; training existing emp1Loyees and recruitment and training of additional employees. Any training required for upgrading responsibilities may best be undertaken in-house by existing employees, but technical assistance from an external source could also be of great value. Against any associated increase in costs will be set the improved revenues already experienced and additional revenue foreseen. Future improvements will require extensions to capital works and increased labor costs. In addition, provision should be made in the estimates for establishing a pilot metering area where ideal conditions are maintained, in order to determine the maximum increase in revenue that may be achieved and any other associated operational improvements. This pilot area should, if possible, be the same as that selected as a pilot waste water district (see Section 5.9). At this stage, estimates of further costs envisaged and budget approval must be obtained. Monitor and evaluate results and budget for further actions. Expand meter testing and repair shop; recruit and train any additional employees required. 4.8 Incentives Some meter readers may have no duties other than to read meters, but frequently they also distribute bills and collect payments. Although it is responsible work, it is normally not well paid. Some countries have found that one of the best incentives is to provide a good uniform which also adds to the public respect for water, and for the water authority. Beyond this, it may be asked whether the meter reader's responsibilities could be increased and her salary improved accordingly. It bears repeating that the meter reader is probably the only employee of the water authority who actually meets the consumer and is therefore a potential public relations spokesperson. She is in the best position to notice waste of water as well as leakage; she could, more easily than anyone else, locate illegal connections. Could she be paid a bonus for reports on wastage, leakage, or illegal connections, and could the bonus be related to the resulting reduction in UFW? Given proper training a meter reader could record readings intelligently in relation to population and usage on a property and note any unusual variations in consumption patterns. She could certainly receive complaints and report them; she might even be able 4-7 to deal with some of them. Could she not also be trained to change defective meters instead of just reoorting them? Would it be cost effective to combine her duties with detailed leak detection? Alternatively, meter readers, inspectors, and fitters might be formed into loose teams for joint action. The word "loose" is important because frequent interchange wotuld help to reduce any meter reading anomalies, inaccuracies, or irregular practices. One other incentive that has been tried is giving discounts on water bills for leaks reported bv the consumer. 4.9 Women Meter Readers It has often been suggested that women should be employed as meter readers. Admittedly there could be problems in many countries (in Muslim countries it would not be possible) but it is a proposal that merits serious consideration. In China, for instance, it would probably be wholly acceptable. After all, meter readers operate during normal working hours, and it is the women of the household whom they can usually expect to meet. Women meter readers may be more acceptable to the consumers. It has been suggested that the particular abilities of women make them more conscientious in such tasks and that they would have the eye for domestic detail which might reveal irregularities in the readings and which might escape the notice of a man. One of the unfortunate traditions is that the meter reader is on a low salary scale -- possibly so low in some countries that dishonesty is encouraged. It may be that, in some countries, the low salary would be sufficiently attractive to recruit women. No doubt it would have to be tried on a small scale and one possibility that could be explored would be to include women as part of a team as suggested in Section 4.8. Any change in meter-reading procedure contemplated following considerations such as deszribed here and in Section 4.8 should be initiated in a pilot area. Such pilot areas should preferably be the same areas selected for investigation of physical losses (see Section 5.9). Examine all possible incentive schemes to encourage employees' enthusiasm for improved metering, while retaining in-house procedure. Consider improved status and possible rewards to meter readers for reporting defective meters, leakage, irregularities, illegal connections, and so on. This activity should be under continual review and could well begin much earlier (see Section 2.2.3). Remember that, because the meter reader is the only employee known to the Ipublic, improved public relations must begin there. Examine the possible employment of women meter readers who are interested: (a) as the major users of water, to ensure its availability at all times and at low production costs by reduction of waste and (b) as keepers of the domestic purse, to ensure equitable payments by all consumers. Select one or more possible pilot metering areas for correlation with pilot waste water district. Pilot area networks must receive 4-8 intensive cleaning, regular flushing, and generally improved maintenance to determine the effect on meter failures. 4.10 Meter Reading by Contract In countries where the provision of public water supply is privatized, underregistration and other domestic metering problems appear to have been largely elimi-ated. There may be political objections to the privatizing of any part of a public service, but where the public service is already failing to produce equitable charges such objections might possibly be overruled. In other countries there may be no such objections, and there such a policy should be considered. Much depends on the terms of the contract. One suggestion is that payment to the contractor should depend on the total quantity of water recorded from meters for which a charge can be made. This does invite a contractor to overrecord consumption and, particularly where a meter has failed, to record an unreasonably high estimate. There is, then, a risk of producing bad public relations that the authority will wish to avoid. Independent spot checks might overcome such objections. In this connection, recent research on the accuracy of small meters has shown that new meters by one manufacturer recorded objectionably high at low flow rates, with the result that some consumers would probably be overcharged. Such factors indicate that management needs to carryout overall inspection of any contract to ensure that it is being implemented fairly. Any contract for which payment is based on results should probably include all aspects of metering, including the remapping of service connections. The selection, installation, maintenance, repairs, and testing of meters would have to be included, along with meter reading and possibLy invoicing as well. If illegal connections have not already been surveyed and eliminated (see Section 3.7) this work could also be included in the contract. In spite of the attendant difficulties, privatization has an appeal which cannot be ignored. A cost/benefit study should be undertaken, for which assistance by a consultant may be required. If studies indicate that privatization is justified, implementation of such a policy should preferably be in stages. The plan could begin with operations in one pilot meter district alone for at least one year. After monitoring and evaluating the results, future policy can be decided. Undertake, in-house or by consultant, a cost/benefit study of privatizing all meter reading (and possibly billing), using an incentive type of contract. The contract may or may not include meter testing and repair, and the survey and remapping of consumers' services, including illegal connections. Subject to the results of the study, implement policy of contract metering by starting in only one district. Monitor and evaluate achievements. Make policy decisions on all further consideration of improvements to metering and any associated billing procedures. 4-9 4.11 Three-Year Prigram The task force must be instructed to prepare and maintain a program for identifying the extent of the nonphysical losses and for reducing these losses to an acceptable figure within a period of two to three years (see Annex 2A). This program will begin and will extend over the same period as that for the reduction of physical losses. Likewise it will be essential to continue excercising control over the ensuring period for the nonphysical losses to be eliminated. 5-1 SECTION 5 EXTENDED ACTION (FIRST FIVE-YEAR PLAN) FOR REDUCING PHYSICAL LOSSES 5.1 Existing Conditions 5.1.1 Organization for Leakage Control: A Reappraisal During the "immediate action" period an attempt should have been made to check basic data as well as possible with the equipment available; to set up a division within the authority to carry out preliminary tests of the system and the equipment; to develop a program for further action based on preliminary economic analysis; and to develop an overall future strategy for additional control of UFW. If the strategy is aimed at further pursuit of leakage control, the first step may be a more detailed review of present conditions prior to taking appropriate actions. Following the completion of the initial investigations and immediate works, there should be more confidence in: the accuracy of measurements of the quantities of water produced and distributed; the percentage inaccuracy of consumers' meters; and the revised estimates of actual consumption for clomestic, commercial, and industrial purposes. These figures will have revealed the extent of nonphysical losses and, by deduction, of physical losses as well. Policy has been determined as to how much additional effort will be exerted initially in reducing "physical loss" by improved detection methods and repairs of leaks. Assuming there has been a "passive" leak detection program in the past -- that is, locating, pinpointing, and subsequent repair of leaks reported by the public --- it now has to be decided whether the existing personnel, equipment, and organizational arrangements should be retained and expanded or supplanted, to comply with future policy needs. It is likely that existing employees will have some, and possibly a great deal of, experience in detecting leaks, combined with a wide knowledge of the system. Often the inspectors have some equipment for locating mains and valves, as well as simple listening sticks, stethoscopes, geophones, or other electronic amplifying equipment; they may also have a modern correlator. They may have some, or a great deal of, experience and knowledge in locating leaks by use of the equipment. (Sounding techniques, insofar as it is possible to describe them, are discussed in Annex 3B.) For any effective improvement to be made in leakage control, employees must be well screened. Those selected for retention should be provided with some incentive, such as an appropriate uniform, which is of much assistance for the necessary entry on to private premises. A review of the current repair organization is required at this stage. This review should cover not only leak detection, but also the repair situation, including deLays between detection and repair and details of any current backlog in repairs. 5-2 It has been stated that, with the introduction of an improved detection process, demand for repairs will increase. But the immediate backlog will be relatively short-lived and, once it has been dealt with, the existing repair staff will probably be able to handle any problems. This is not necessarily true, however, because if physical losses (measured as a percentage of net production) appear to have been fairly steady over recent years then, as the system has expanded the actual number of unrepaired leaks has increased. In such conditions the repair staff has not been adequate. With the introduction of improved detection, then, a permanent increase in repair staff will probably be required. But the short-term need may best be met by employment of contract labor, as long as adequate supervision is provided and good standards are ensured. Sufficient plant must also be provided. At this point, a preliminary estimate should be made of the probable labor requirements for dealing with the increased volume of leakage likely to be found. Review existing procedures for leak detection reporting and repairs. Consider all possible improvements using only existing employees, including more intensive visual inspection particularly in dry conditions and in periods when pressures are high (see also Annex 3). Determine the extent of delay in repairing reported leaks and take action to reduce any backlog including, if necessary, temporary additional contract or other labor. Review existing equipment and assess needs. Intensify repair organization as necessary to deal with extra leaks detected. This may be needed only temporarily, until the backlog of leaks found disappears. 5.2 Mapping, Records, and Network Analysis (See also Section 3.8.4) Control of UFW entails the continual repetition of a series of simple, logical processes and tasks to obtain increasingly accurate detailed data that facilitate ever-more-efficient detection of deficiencies. Mapping is no exception; it is a continuing process required for meeting the needs of system expansion and new demand. Maps and records should already have been given some attention, and hopefully some updating and improvement has begun. These preliminary improvements can now be examined, and decisions can be made regarding further improvements required, particularly regarding how much can be done in-house and the extent to which specialized contractors should be brought in for assistance (to locate buried valves, map all mains and services, and so on). The future computerization of mapping should be kept in mind for any contract work. In any case, mapping should include elevations of all valve and hydrant covers and other key points for future computer network analysis. Many records may have been kept that apparently have no permanent value. Yet with a little expansion or modification they may be useful. For instance, a continuous record of leaks detected and details of repairs carried out will show the existing time delay, although otherwise such records may 5-3 never be used. If these reports were cross-referenced by street, a valuable record would be built ui to show where leaks are concentrated and the cost of individual repairs. Such data would be useful in determining when complete refurbishment of a section of the system would be justified. Computerization of records would, of course, make data retrieval simpler. It is reported that in Holland costly individual repairs may be replaced by computerized data retrieval followed by wholesale system replacement. With the completion of district metering and the monitoring of results, little more can be done to improve the leak detection process without accurate mapping of the network. The mapping that is required is not just geographical. For logical leak detection, knowledge is required of the material and age of all pipework, valves, and fittings. This information must be built up slowly by studying old records. Such information is also required as part of the normal operation of the undertaking and is invaluable for developing computer network analysis. A network analysis may be required at this stage for other purposes, such as economical network reinf-orcement. Remapping could also be of enormous benefit for modifying zone or district boundaries. Review progress on whatever remapping of the distribution network is necessary, beginning in the areas where the greatest leakage and metering problems are expected. Remapping should include finding, exposing, marking, and numbering all valves. This work may best be done in conjunction with remapping of consumers' service connections. Consider having all remapping completed by private contractor (covrering also the possibility of computer mapping in the future). This work should be done at the same time as the work outlined in Sections 2.3.6 to 2.3.11. 5.3 Future Training, Labor, and Equipment Where a program of intensified leak detection is proposed, training with ai authority in a developed country is invaluable, even when employees already have some experience. Training centers exist in many developed countries where collaboration with water authorities will provide intensive practical training covering a wide range of equipment and methods (see also Section 3.8.3.iii). In the first instance, training might be limited to three or four carefully selected employees who will be directly engaged in leak detection and one engineer or technician. The latter can then play a key role in organizing the intensified service and recruiting and training additional inspectors. If this is not feasible, an alternative might be to engage some other authority or a specialized consultant/contractor to undertake all additional training "on the spot," providing any additional equipment required as part of the training. The training might also be related to a contract for the first phase of leak detection., Such training should facilitate an assessment of the additional labor and equipment required to fulfill short-term, and possibly long-term, 5-4 policy. Suitable equipment is manufactured by a number of the developed countries under various trade names. Most LDCs have technical connections with one or more developed countries and can obtain details of equipment available from the consulates. Arranging for such training is one of the first steps to be taken in implementing a program for improved leak control. Arrange for additional training in leak detection of employees selected from existing leak inspectors as well as one engineer/ technician suitable for providing future extended training. This training may be on the spot, but initially should include training with a foreign water authority. (See also Annex 5A.) Purchase additional equipment considered essential. Establish stores inventory for all tools and equipment. 5.4 Technical Assistance When an authority is unable to devote the time of its own staff to a detailed study of the economic justification for and the establishment and operation of an active leak control program, there is a clear need for assistance by foreign consultants (which may include a foreign water authority). When this is necessary, it is essential for the associated contract to include a major item for training an adequate number of the local staff to continue the program. It must be stressed that leak control is not a short-term actvity which once begun can then come to an end. Experience shows that unless the control program is perpetuated indefinitely, soon after cessation the UFW figure is as high as ever. The need for appointing a competent, long-term consultant to assist the authority should be seriously considered. A key part of any contract with a consultant is the establishment of an effective permanent organization in the authority and a commitment to continue advising the authority on the detailed solutions of problems encountered during the first year or longer. Of equal importance is provision by the authority of a sufficient number of suitably motivated, intelligent, and qualified "counterpart" employees to undertake on-the-spot training and take over the operation at the end of the consultant's contract. The organization must be established either before or at the very beginning of the consultant's contract, if it is to be fully effective. There are far too many consultant's reports scattered around the world recommending institutional changes, personnel training, purchase of equipment, and overall programs for improved leak control that have never been implemented and are at best coated with dust -- or at worst, lost. In order to avoid this it is often better for an authority to set up an organization with the help only of an individual consultant for initial advice and subsequent follow up as necessary to help renew enthusiasm and initiative as well as to overcome difficulties. 5-5 If considered desirable, apply for and obtain a very brief preliminary study and advice from external source such as (a) foreign water authority, (b) individual expert, or (c) Bank or other technical assistance. This study is to assess short-term needs and to advise on iaction program from time to time. 5.5 Pressure Reduction (see also Section 6.3) During the immediate action program there should have been some preliminary tests of the effect of pressure reduction on leakage. It is only recently that the impact of increasing pressures has been fully appreciated. Previously it was assumed that the leakage flow would follow the orifice formula, in which the flow is proportional to the square root of the pressure. Evidence is now clear that in most areas the pressure flow relationship curve does not follow this formula; instead, it normally curves in the opposite direction, which is much more serious. This phenomenon appears to result from the flexing of splits and similar openings so that the orifice size increases with pressure. The emphasis is now placed very firmly on effective control of leakage by making all practicable reductions in pressure (see Figure 3-3). This may be achieved by means such as installing pressure reducing valves (PRV's); utilizing break pressure tanks to increase the number of different pressure zones; reinforcing the network to minimize friction losses and permit operation with reduced working pressures; changing the service connections to mains operating at lower pressure; or introducing speed control or other means of reducing night flows from pumps. PRV's may be installed in a leading main adjusted to be fully open and to provide no reduction in pressure at times of peak flow but to operate as flows are reduced, thereby preventing the build-up of pressures during the night. Better control is achieved by means of telemetry-controlled PRVs. In some cases, zone boundaries may be changed to reduce excessive pressures. Every possible means must be considered for effecting reductions in pressure, but unless the effectiveness of any change can be accurately measured, there will be no way of making a cost/benefit study of any proposal. Measures adopted must therefore follow accurate pilot studies or temporary tests introduced by manual valve throttling or other means. There is, of course, a strict limit to the number of changes that can be justified in economic terms. Regular monitoring of pressures in all parts of the district should be instituted so that the results of pressure modifications may be evaluated. In the first place, any tests undertaken during the Immediate Works program for measurement of pressures and night flows in a pilot area may be extended to other areas. Make every effort to reduce all excessive pressures in the distribution network at all times of day and night. First reductions should be by sielecting and isolating a pilot area, and monitoring the effect by meter. Pressures and resulting leakage frequently increase at night, and reductions in night pressure may be achieved by simply installing in-line pressure reducing valves (PRVs) without 5-6 significantly reducing the daytime pressure necessary to ensure continuous supplies. A careful survey of fringe areas to zones and reconnecting of service connections can be rewarding. Interzonal leakage must be strictly controlled. 5.6 Monitoring District Flows An initial check on net production should have been made as part of the immediate action program. Further checks are required not only of net production for the whole system, but also of the separate zone and d-istrict flows, at regular intervals at least four times a year and preferably much more frequently. The initial flow figures were, at best, estimated; but every effort must now be made to install meters to record or at least indicate flows over a 24-hour period into each zone or district. When a district is first established, all waste detection should be concentrated in that district and consumers' meters should be checked in order to determine the bases for demand and consumption. As conditions improve over the years, these bases may be lowered. But they are needed at this point for comparison with the flows recorded each time the district demand is monitored. Regular monitoring of district flows is the simplest method of locating leakage areas and, if inspectors are concentrated in those areas, of effectively controlling leakage. Even where an authority has introduced waste meter districts over the whole system, there are times when inspectors are not able to monitor the "waste" districts. In such cases, control can be effectively exercised through evaluation of district meter flows. One important method of improving control, particularly where waste-water districts cannot be operated, is to gradually break down the main district into smaller districts, each of which can be metered. Where district meters have not yet been installed, they should now be regarded as essential. Once installed, the meters must be tested for accuracy at least once a year. Any necessary improvements should be made in metering, in order to facilitate regular monitoring of flows (see Annex 5B). Isolation valves between zones should be checked regularly to ensure that they are not "letting by" at any time. Valves between districts should not be "letting by" during flow tests. It is likely that this will entail considerable work on valve reconditioning -- or scrapping and replacement. Pressure checks both inside and outside the boundary are necessary to test valve integrity and to ensure that there are no unknown cross-connections. District meters must be of a size suitable for measuring the whole range of flows with reasonable accuracy. They are normally installed on a bypass but, where this is impracticable, turbine meters inserted under pressure are very valuable. They should be integrating and recording as well as indicating, unless it is found that standards of maintenance are such that the simpler indicating meters are more reliable. A common size meter for small districts is 6 inches; for larger districts, two or more entry points with 6 inch meters may be better than using a larger meter. The presence of large private tanks and indirect plumbing may limit accurate testing to low demand periods, when flows can be monitored to give 5-7 meaningful results of net night flows. But tests made at times of high demand can also provide valuable information and should be undertaken when feasible. Check the 24-hour variation in production for the whole system and, as it becomes increasingly possible, for separate districts. (See also Annex 5B.) This normally has to be done during a period when demand is low, and when it may be expected that all private tanks will have been completely filled during the earlier part of the night. 5.7 Minimum Night Flows Data on inflows to each zone or district are of inherent value since variations in these flows over successive periods will help to reveal hidden losses. But these gross night flows cannot be strictly correlated with losses, because they will almost certainly include night consumption by industry or public services, such as hospitals. The aim is to determine the night-flow rates, if possible including variations throughout the night, for all the large meters in each district. Normally, where there is legitimate night consumption it is possible to arrange for this to be recorded by the consumer who, if properly approached, will appreciate efforts to reveal possible leaks or waste in his network. -jnmetered night consumption by public lavatories, street cleaning, and so forth must also be estimated. Having determined the figures for nondomestic consumption during the night and deducted this figure from the gross inflows for each district and for the system as a whole, the net night flow can be determined and expressed as a percentage of average demand or in liters per property per hour. This figure will vary for different districts and for different authorities depending on a number of factors, including the average night pressures in the areas where leaks are prevalent. It should not be greater than 25 percent of average in any system. It should be possible to reduce the figure below six liters per property per hour (4.3 cubic meters/property/month) by intensive waste control measures, but the target in an old system is unlikely to be less than ten liters per property per hour. The minimum net night flow is likely to be the best indicator of whether or not there is excessive leakage. Retest production meters (or estimates). Repair or install additional meters as necessary. Install additional meters at all treated water and source pump stations, service reservoirs (inlets and outlets), and boosters. Test for leakage in (a) service reservoirs and (b) transmission mains. Begin regular monitoring of production and district flows. Take sample readings on all large consumers' meters to assess industrial and commercial night consumption for the whole system. Adjust the total night production flows to determine the net (domestic) night demand and express as a percentage of production. This figure should not exceed 25 percent (or 10 liters per property per hour). Repeat for each district. 5-8 5.8 Inspectors for Valve and Hydrant Operation The inability to operate valves effectively is probably the most difficult problem to be overcome when introducing any form of active leak detection. Frequently this stems from low standards of construction for the initial main laying and the perpetuation of low standards in the repair works following mains fractures. Great care must be exercised to prevent the entry of stones or, during repairs, to remove stones and debris from inside pipelines. All too often one sees main laying or repairs in progress with no attempt at protection of open ends and no subsequent high rate flushing out of the system through washouts. When stones and grit are not removed, they inevitably become trapped in the valve seatings and prevent effective closure. If valves are operated carelessly, with excessive force, and no attempt is made to dislodge grit by repeated careful opening and closing against high flow discharges, the usual result is a valve with a bent or broken spindle, which is quite inoperable. Similarly, the attempted operation of a larger valve by one person produces an eccentric force and, again, a bent spindle. Another problem arises from lack of care and discipline in the operation of valves. When valves are shut off for repairs, closing one valve on each side of the repair will normally suffice. But if they will not close effectively, then more remote valves must be closed until the flow can be shut off. The problem then is to remember, when the repair is finished, to open all the valves that have been shut off. This problem is compounded when some of the valves in the system close clockwise and some counter-clockwise, since it becomes difficult to determine which valves have been reopened. Unless rigid discipline is exercised many valves in a system are left closed, with the result that flows and pressures are reduced, supplies fail, and the quality of the water in "dead" lengths of main deteriorates. A further difficulty arises when, as a result of road resurfacing, valve covers are broken or valves are covered over and lost. Frequently there are no valve marker posts to assist in relocation. Similar problems occur with buried hydrants. These problems usually occur when the responsibility for valve and hydrant operation and maintenance is not clear. Experience shows that it is essential that all such operations be undertaken by a limited number of experienced and responsible employees, who may be graded as senior or "district" inspectors. Experienced, older employees should be trained as senior inspectors and given the sole responsibility for operating and maintening valves and hydrants. No repair gangs should have valve keys among their tools. A senior inspector's main responsibility should be to personally ensure that valves are operated correctly, with labor assistance as necessary. They should also ensure that valves are maintained in proper working order, complete with marker posts, and are repaired or replaced when found to be leaking or in any way defective. An inspector should ensure that valves are not damaged or lost 5-9 when road works are in progress. Finally, inspectors must protect the mains generally against damage by operations of other utility workers or during road reconstruction. These inspectors are likely to be more than fully occupied. Yet at certain periods they may have time to undertake leak detection in the area they are controlling. Before active leak detection can proceed very far, considerable effort will probably be required to find, map, mark, and change all defective valves, beginning with those required to isolate zones and districts effectively. This may prove to be a mammoth task, but the effort required will be amply justified by the improved operation and control of the system. In some systems the number of valves may not be sufficient to exercise proper control. It may also be found that there are insufficient washout valves or hydrants to permit flushing of the system, either on a regular basis or following shutdowns for the repair of leaks. In fact, one of the serious problems for leak inspections in some systems is the absence of any direct contact points. Valves are buried; hydrants are sparse; and there are no stop taps on consumers' services except those adjacent to meters situated on private premises not accessible at night. The first essential step in such cases is to restore, and if possible expand, the installation of valves and hydrants. Select top grade district inspectors to have the sole responsibility for all maintenance and operation of valves, including initial location and subsequent guarding of surface boxes (see Section 2.4.5). During the same period install district (or zone) meters. Ensure that districts can be isolated permanently or, if this is not operationally possible, at least temporarily for test purposes. Concentrate lieak detection and repair in districts that appear to have the most leakage. Always measure and record the quantity of leakage detected and stopped. 5.9 Pilot Waste Meter Districts In developed countries the monitoring of district flows and the consequent concentration of active leak detection in areas with high night flows has been shown to be economically justified, where the marginal cost of water exceeds about $0.022 per cubic meter. This figure is likely to be very different in developing countries, depending on the cost of labor and production. For any developing country authority, the economic assessment of the cost of active leak detection in relation to the benefit in marginal cost of water saved -- an assessment based on preliminary estimates of net production flows and actual consurnption -- has probably justified the provision of district meters and the monitoring of district flows. As the regular testing of meters and monitoring of flows proceeds, the extent of physical losses in each district can be estimated with ever greater accuracy. At this stage, certain information is required to justify 5-10 the breaking down of each district into a number of waste-meter districts in order to further concentrate leak detection activity. This information can be gathered by establishing a pilot waste-meter district. The pilot waste-meter district should be located in an area where there is considerable leakage. For test purposes, it must be isolated from the rest of the network; thus it is essential that all isolating valves be operating effectively. The area chosen should, preferably, be one of the areas selected for accurate measurement of actual consumption (see Section 4.10). It may be found that to ensure effective valve operation the simplest course is to replace all valves in and around the pilot area. In the process, care should be taken to flush out all the mains thoroughly in order to remove grit and stones. The area chosen should be operating under a high working pressure at all times and should serve about 1,000 properties. If the area includes a number of nondomestic meters, the night flows through these meters should be readily determinable. Other factors for consideration are demographic Levels, ages of mains and services, and previous history of leaks. The aim is to be able to isolate the pilot area and to supply it through only one connection. To accomplish this, a waste meter of size appropriate to the maximum and minimum flow ratio and a PRV should be installed on a bypass. The rate of flow into the area must be measured over a 24-hour period. By comparing the flow with the total metered consumption, total physical loss can be determined. As an exercise, this flow can be repeated with the PRV set and maintained at varying lower pressures in order to determine the effect on unmetered physical losses, and consumption. The pilot area should then be subjected to nighttime "step tests" to determine the night flow to each street, or sub-area. (For details on step tests, see Annex 5C.) This will locate those parts of the area producing the greatest leakage and enable leak detection to be concentrated in those limited areas. Repeated tests, leak detection, and repairs will demonstrate how much water has been saved and indicate both the threshold for acceptable leakage and the cost of detection and repair. This investigation will be invaluable in determining the extent to which other waste-meter districts should be established in order to optimize the benefits from improved leak detection. Where operating pressures are very low, however, waste meter districts are not practicable. Select one or more pilot areas and install a waste meter, making sure that all valves can be effectively closed. Undertake regular tests of 24-hour and night flows and, if possible, regular step tests. (See Annex 5C.) Initially, test monthly for training purposes; later every six months. 5-11 5.10 Night Domestic Consumption In Section 5.6, reference was made to the variations found for net night flow in different systems and in different districts of the same system. When domestic consumption is not metered, it is not easy to proceed further in investigating the variations. Although the differences have to be accepted, pilot tests on selected waste meter districts can enable an authority to arrive at reasonably cautious targets for the future. Where domestic consumption is metered, providing the accuracy of the meters is known, the variations in average per capita consumption between districts can be determined. This figure can then be related to net night consumption in order to arrive at a reasonable target. To make this calculation, some boun,dary adjustment of meter reading records, or "walks," may be necessary. These comparisons will help also to reveal possible sampling on other errors in arriving at meter inaccuracies. They will also facilitate a reassessment of the proportions of UFW attributable to physical and nonphysical losses. Study the 24-hour total production of water in relation to average total metered consumption in each district. For this purpose, boundary adjUstments may have to be made to meter consumption records. 5.11 Waste Wasted water, or simply waste, is an all-embracing term that includes leakage. In countries where domestic water is not metered, waste -- whether the willful failure to turn off taps or drips due to defective fittings -- is normally regarded in the same light as water leaking underground from fractured or corroded pipes and joints. Physical action can be taken by the authority to reduce the latter but the former can be prevented only by (a) the strict application by bylaws designed to prevent the use of bad workmanship and fittings and (b) by public support, through the proper operation and maintenance of private fittings. The waste inspector is responsible for preventing waste in every sense, which includes supplying and fitting approved equipment and checking that it is being used properly, as well as finding leaks. Some authorities provide free re-washering of cold taps in order to improve inspection and to educate the public in the prevention of waste. In countries where domestic supplies are metered, there has been little concern over waste other than leakage. The view has been that consumer waste is metered and serves to increase the revenue. This view is probably not justified because, whereas normal consumption occurs only for brief periods, waste from fittings as well as leaks on private service pipes may involve only small flows, they are continuous and thus may amount to a significant volume. These small flow rates will almost certainly not be registered on any meters, except during periods of actual consumption. They are therefore of great importance and merit the same close attention as underground leaks from the authority's own system. Recent surveys in developing countries have revealed as high a percentage of defects and leakage in private pipework, even where metered, as on public mains and services. Notices to property owners to repair leaks are often ineffective. 5-12 It has been supposed that when an authority has completed a project and has surplus water, waste by consumers is actually of financial benefit to the authority. But this is only true if it is recorded on the meter. In many developing countries it will be found that: bylaws, if they exist at all, are honored only in the breach; there is no compulsory control or registration of plumbers to ensure a reasonable standard of workmanship or limitation on illegal practices; there are no effective standards for private plumbing fittings; and overflows and drips go unnoticed. Some authorities are better organized and have insisted that all overflows discharge outside the building -- preferably into the public street. Normally, however, such restrictions are confined to authorities in countries where domestic supplies are not metered and where the need for waste control is better appreciated. Where bylaws exist, they have to be put into practice. After publicizing, repeatedly checking, and serving notices, it may be necessary to resort to publicized prosecution. Where bylaws do not exist, the attention of the central government of the country must be brought to bear on this issue. (Typical bylaws, not generally applicable without modification, are provided in Annex 5D.) The cooperation of the public is probably more valuable than the strict application of bylaws, but both will only be established over a long period by education and the gradual improvement of social standards. A public education program should be considered by management. Inspectors must be trained to appreciate that their work includes discovering waste as well as leakage. Re-estimate the proportions of UFW due to metering deficiencies and waste. Review problems of water waste, both public and private, and determine action to be taken by improved education and public relations as well as by strict application of bylaws. Review existing bylaws and make changes as needed. Where bylaws do not exist, prepare legislation for approval by government. (See Annex 5D.) Experience shows that most leakage problems are due to innumerable small leaks on service connections. Even when domestic consumption is metered, small leaks can often be found on consumers' pipework. Correcting this problem is usually the responsibility of the householder, upon whom notice should be served to take action within a certain number of days. If no action is taken, faults should be rectified at the householder's expense. Repairs are a time-consuming and tedious task. Concentrate the leak detection and repairs in districts where night line tests suggest the greatest leakage losses occur. Do not expect that correlators or other modern equipment will normally solve the problem, though they may be invaluable for locating certain leaks. Monitor and evaluate for a trial one-or-two- year period. Pinpoint possible areas where complete renewal of a network may be the most economical solution. 5-13 Make an economic study of the reduction already made in leakage achieved thus far. If this part of the UFW (that is, excluding an estimated amount for underregistration of consumers' meters) is still greater than 12 percent, review the procedure to determine (a) possible improvements in control without changing the basic control method; (b) additional potential savings by introducing waste meter districts or by complete replacement of all mains and services in selected streets (see also Annexe 5B). Re-examine policy and budget. 5-14 Annex 5A. Training Most large authorities in developing countries will have a few employees who are engaged in locating the source of leaks. They frequently have some equipment for locating pipes and valves, and they should be familiar with the sounding process referred to in Annex 3B. Most of their technique will have been picked up gradually and may, or may not, be effective. Their equipment is often antiquated but may still be effective; it may have been updated and may even include a modern correlator. Up-to-date training may already have been provided by suppliers of equipment. More often than not their experience has been limited, their equipment is outdated, funds have not been provided for improved performance, and most importantly, there has been little interest or enthusiasm from management. The employees have lacked discipline or control and, without these, have suffered from boredom and frustration. To transform this passive control into some form of active control, a new spirit of enthusiasm--backed by funds--has to be implemented. The best start is to arrange for about three selected employees to be transferred for a period of up to ten weeks to a developed country where they will be given special training using modern equipment and will work with trained inspectors employed by another water authority. If two or more employees working in the area of repairs could under go similar training, it will be advantageous to the Authority. They can learn to appreciate the standards of workmanship required to ensure the best possible repair under difficult conditions. If possible, one of the employees selected should be an engineer or technician with an aptitude for teaching. She should then undertake the selection and some training of additional employees required to build up the organization in the area of effective leakage and waste control. This initial training of a few selected employees should be part of an ongoing program. As the division expands, one of the incentives for new recruits should be the potential to be selected for training abroad. Overseas training should normally provide short, formal courses at specialized training centers that include: (a) theory of leakage control; (b) practical demonstrations and experience in the use of a wide variety of equipment; and (c) maintenance of and simple repairs to all equipment. This should be followed by a lengthy period of hands-on operational work during the daytime and at night, covering all forms of active control including recordkeeping and analysis of results. This hands-on experience should include electronic detection of pipelines and cables, experiments on measurement of waste under different pressures, and flow measurements. It should also include tapping mains under pressure, insertion of pitot and 5-15 turbine flow meters, repairs to pipes and services, flushing mains and sterilization, testing valves to ensure tight shut-off, the use of portable test flow meters, pressure gauges and recorders, and setting up a district metering system and a waste meter district. 5-16 Annex 5B. District Metering for Control of UFW When the manager of a water authority is satisfied as to the total volume of water being put into his distribution network, his next aim should be to learn more about how this water is being distributed. The overall objective is to supply the water as economically as possible. To fulfill this goal, the manager and his staff must have detailed knowledge of how the system performs. The quickest way to obtain this knowledge is to divide the network into separate "districts" of a convenient size such that the flow to each district is permanently (or temporarily for test purposes only) supplied through one or more meters. These districts may be separate pressure zones or may serve separate areas at the same pressure. The essential point is that the district can be completely isolated, except at the metered entry points. Ideally the district should include 5,000 to 10,000 properties but, if the authority has no previous experience with districts, it may be more convenient to begin with larger districts and then, as knowledge and control of the supply improves, to break these down to smaller districts. It is simplest if the supply to each district is through only one meter, but if this is not practicable, the use of two or three meters is acceptable. Where the network is based on a ring main supply, the permanent isolation of districts may not be easy. The meters must be accurate, so they must be tested frequently using a portable test meter fitted to a hydrant. If possible, they should be installed on a bypass, but if they have to be installed on a main conveyor that cannot be easily shut down for the necessary installation of valves and fittings, the alternative is to install either a pitot meter coupled to a transducer (for accurate reading of low flows) or else a small turbine meter. Both of these meters can be inserted under pressure without shutting off the supply. Meters fitted on a bypass are normally the 6" or 8" diameter turbine type or larger, as necessary to measure the maximum flow without excessive loss of pressure. They must not be so large that low night flows are not measured accurately. Regular monitoring will show management how the system is working, what changes are taking place, and what actions have to be taken to meet new demand. For purposes of UFW control, the aim of this procedure is to monitor regularly the flow to each district over a 24-hour period. If this is not possible, then monitoring can take place during the night hours only. The net supply of water must take into account the gain or loss of water in any service reservoir within the district. It goes without saying that structures such as reservoirs, boosters, and PRVs must be operating satisfactorily, and management must be aware of any reservoir leakage. The district should first be monitored quarterly, then monthly, and later weekly; only exceptionally is more frequent monitoring required. The records will consist of meter reading books or cards showing variations over 24 hours, or at night only, at monthly or weekly intervals. Graphs of meter readings should be kept to highlight variations. Nondomestic use in each district can be recorded separately and any variations in domestic 5-17 use, which will include all leakage, will be evident. Domestic use could be translated to liters per property per hour, and the figure for each district compared. Any anomalies can be investigated and, if caused by excessive leakage, will serve to concentrate leak detection activities. Of particular importance in this comparison will be the average pressure in each district. It must be expected that leakage losses will be higher in those districts with higher main pressures. Pressure variations in each district have to be determined and regularly monitored using pressure recorders installed at strategic points. A lower than expected night pressure is often an indication of extensive leakages. Having established a district, every possible step should be taken to locate and repair all leakage in that district in order to determine night flow threshold conditions. These may be varied later as monitoring is continued, district comparisons are made, and records are built up. The regular monitoring of districts and the breaking down of the system into smaller and smaller districts may be found more cost effective than the introduction of small waste meter districts with frequent night step- tests. 5-18 Annex 5C. Step-Testing in Waste Meter Districts Before carrying out step-tests on a waste district, it is essential to go through a checklist to ensure that: (a) a map of this district has been prepared and is correct (see Figure 5C-l); (b) a program of the procedure has been prepared, complete with timing instructions for valve operation; (c) each valve shown on the plan can be found, is operable, is correctly set, and can be closed "drop-tight"; (d) all large meter locations are known and, where necessary, consumers have been notified of shut-down; (e) the police and fire service have been notified; (f) all supplies for hospitals and essential services are maintained; (g) all employees have the necessary maps, instructions and equipment including flashlights, watches, valve keys, cover keys, pressure gauges, fluorescent jackets, and two-way radios; (h) the waste meter clock is correct and all watches are time- checked. Incidental equipment for normal repair operations should be available on the spot. It is essential that any new program be well rehearsed in daylight before proceeding with a night test. Security of personnel should be given special attention. Typical records of waste meter district tests are given in Figures 5C-2 to 5C-8. There are various types of step-tests, such as the three noted below: (a) Christmas Tree - This is the standard method extended to give a reverse opening sequence. This extension may not be practical in situations where low night flows are of very short duration. (b) Open and Close - This is a method where each section is turned off and on again for a brief period so that the supply is interrupted for only a short time. The results are much more difficult to analyze but the procedure may be necessary to better maintain supplies for essential services. (c) Double Locking - For this test, as the metered supply is cut off to a section an alternative supply is made available from outside the district. This is ideal where it can be arranged but requires more labor and an assurance that all valves, not only the boundary valves, close drop-tight. 5-19 When conducting such tests it may be desirable to have in mind that: it is impracticable to find and repair all leaks; it may be uneconomical to repair Large numbers of small leaks. Figures for acceptable leakage vary but generally in old districts in the United Kingdom flows of up to four liters per property per hour are acceptable, whereas in newer districts lower figures apply. In many old urban areas, when beginning to practice active leak detection in this way, it will probably be difficult to achieve minimum net night flows less than 10 liters per property per day unless the average pressure is relatively low. If operating pressures are too low, step-testing will probably not be a practical method of leak control. Experience shows that in many districts where waste water meter districts have been established it is more cost effective to treat the meters in the same way as district meters; that is, to monitor the 24-hour flows frequently, evaluate changes that occur, and concentrate leak detection accordingly. Further concentration of activities to single streets is often not justified especially where labor is plentiful and not expensive. 5-20 Figure 5C-1. PLAN OF WASTE DISTRICT 4 - 65 Everst Road C elore Road 3 \4523 - 42 48NCorsor Clos20/ EnsIn Way Cloe 27 6" 25ir Clse 44rs renlneA 1 38^~~39 \ 47 J 2 \ \~~~~~Brittania Way @, g WASTE METER \ \ ~Hannibal Rdw Hodnan Way \ 4" \ / 4SLngfod Wy 4 \' \ @) t 53 2 4 n Cordelia RoLd 49 3~~~~~~~~~~~~~~~~~~~~~4 Valve Nt O drobisher --- 7 /alvCres Cs54 CambriA Gdns/Ck t Qk t / old/ /Cordelia Gdns Hannibal 55 /3 "' 'Rd 5 3 6 \f)2 Cg)\ an pus Way / E liza bet han Way / ordel ia Road / 66," 4# \66 /4Clare Rood \ 68 a / Xho~~~~~~~drian Way /C6ra~tnford Ave. \ 95~~8 / ; \ Elsinore Avel/| Caroc v @ 7~ ~ ~~3 6" \Cledonia Rd 3L Osborne Ave / 61)vj5~~icount Road \ /|6 _ / \g) / / Hi~~~~~~~28llingdjon Ave 4 Clare Road 29 26 / ( 3 3"~~~~1 30,7 Explorer Ave 2- 18 6 2 34 33Cye\odMulberr Ave I \ ~~~~~~~~~~~~~~~~~Longford Way 4" / ; 32 \ ~~~~~~~~~~~~~ ~ ~ ~~Longford Ave ma 6 Long Lane Town Lane,.. I& Boundary Valve Hollywell W ay 14 13 4/ 0 Circulating Valve16 _/ v Step Valve 16/ 25- i Step No./ s Valve Not Operoted 14 Valve Not lncluded In Programme V'alve Closes Anti-Clockwise World Bank-30505 '16 5-21 FiguLre 5C-2. Typical Step Test Chart District No. 2 Street, Remarks: Date: on Wednesday February 27, 1985 off:Wednesday February 27, 1985 .0 III POW I X 'or m ,,000 0 2100,000 \ \ \ \ \ \ \7\ 0~ 500 0 2 Worl CO n-350:1 C~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I 10000 __ _ _ 0~~~~~~~~~~~~~~~~~~~~~~~~~~~0 0~~~~~~~~~~ 0~~~~~~~~~~ 30,000 0 ______~~~~~~~C4 0 1 2~~~~~~~0 I ,30,000 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~ol ok-00: Figure 5C-3. Meter Reading 'epor NON-DOMESTIC METER READINGS: _ ____ WASTE METER DISTRICT NO:__ Initial Reading Second Reading Average Walk Serial Meter Hourly Number Number Size Consumer Meter Location Date Time Reading Date Time Reading Consumption Consumption Worid Baink -30505 14 5-23 Figure 5C-4. Leakage Report WASTE DEPARTMENT DEFECT SUMMARY WEEK-ENDiNG Waste Date ot Works District Works Instruction Nature of Leakage Number Instruction Address or Location Number or Defect Leakages/Defects Reported During Week = =______ Leakages/Defects Repaired During Week Leakages/Defects Outstanding From Previous Weeks Consumer Side Leakages Reported During Week Consumer Side Leakages Repaired During Week Consumer Side Leakages Outstanding From Previous Weeks World Bank-30505:8 Figure 5C-5. Property Report PROPERTY SUMMARY: (WASTE) DISTRICT NUMBER:_ Un-Metered Properties Metered Properties Number of Length of Public Road/Street Etc. Date Services Main Houses Bungalows Flats Shops Offices Industrial Others Domestic Shops Schools Churches Houses Industrial Others World Bank-30505 13 AVERAGE HOURLY CONSUMPTION (m21hr) t~ ~~~4 C, 0 0, 'o I 1 0 :X 1 ? ne Spring 1 -- - ~~m Summer m … i@_ ___ _ _ _ _ _ _ Autumn 0 Winter 0 Spring …- ~~~~~~~~~~~~~~~~~~~~~~~~Summer Autumn _ Winter Spring Summer = = = = _ =- =-…- - =-= = = = = = = = _ = _ =___ _ _ Autumn n 8@ Winter t= Spring a- __-_…_ _ _ _ ____ _-_…S_ __ _ -_ _ _ 5ummer Autumn -…I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Spring Summer Autumn Winter Spring Smmer --I r ~Autumn m 0 Winter ; Spring Summer Autumn 0 Winter Spring summer Autumn Winter m -4 Spring M> Summer M Autumn r' o Winter0 Spring C/) Summer - Autumn '-' winter z Spring summer Autumn jo Winter Spring Summe 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Autum Figure 5C-7. Whole Area ReLort DIVISION WASTE CONTROL STATISTICS YEAR ENDING-31/311985 WHOLE AREA Month APR MAY JUNE JULY AUG SEP OCT NOV DEC JAN FEB MAR WW\M Areas 2 7 5 6 3 4 4 2 3 Open Runs WWM Areas Step-Tests 1 3 3 1 2 3 2 1 2 1 Total Saving on WWM Areas 6000 9893 3240 18,350 17 025 6000 57.600 Litres/hour Calculated Daily Saving 0.120 0 198 0 065 0 367 0 340 0.120 1 1b2 Mil/d Average Daily Supply tor the 115 849 117 211 122 753 11°533 113 702 111 696 107 236 106 819 107411 112 147 111 354 112 810 Month MIl/d Savings as a % of Average 0 11 0 17 0.06 0 32 0 30 0 11 1 02 Daily Supply Total Water Supplicd 1984/1985 41 322 969 Ml Total Savings 1984/1985 110 108 I!h Average Daily Supply 1984/1985 113 214 MI/d Calculated Daily Savings 2 362 MI/d %of Total Water Supplied 2 09% World Bar,k 30505 7 5-27 Figure 5C-8. Night Flow Report GRAPHIC RECORD OF NIGHT FLOW (WASTE) DISTRICT NO._ MINIMUM NIGHT FLOW Date of Open Reading Scale in ml 100…-… -90-… -80-… 70… -60… 50-- _ E - -40…- - - - - -3C0-…--- - 20-…- - - - - NET NIGHT FLOW Date of Open Reading Scale in ml *100 … -90…- -80-… 70-… 60 -i I; 1 1 40… 30…- 20-…- 10-…- World Bark-30505 9 Figure 5C-9 Step-Test Summary: Waste District Number: 1 Road Cloe A Drp Properties Litresi Date Totol SteDVlv C Recorded Propertyl Metered T,ime Consumption Number Locathon ot Valve (da±k) Litres Unmetered Metered Hour Consumption Remarks 4-9-1984 STAR1 On Pavement 02-52 5,400 1 aS _______ Rd A 100 Used As 2 On Pavenent A CV Ave A - 02-56 5400 3 On Pavement A NIL 03-00 4,800 4 On Pavement (Ashtray) A 600 03-00 4OO ~ _______________________End Road A 60 03-05 4,600 5 In Roodway Rd Rd A 200 03-09 4.450 6 In Roodway Junction oa A 150 _ _ Lan_ Le _ _ _ Rd 03-13 4.400 7 In Roadway PO A 50 03-t8 4.200 8 In Rcodway Opposite Rd A 200 03-21 3.850 9 In Roodwayr JunCtIon Of A 350 Rd _ ._ Lane 03-24 3,750 10 On Footpath at Rivet Rrldge A 400 n ______._______ Rd 03-28 3 750 11 On Pavement Opposite C NIL _ ~~ ~~~~~~~~~Rd _ 03 33 3,750 t2 In Roadway OIS Kay's Corner A 350 _ _ _________________ Rd A __ 3 _ 03-37 3,800 13 In Roadway. Junction ot A 600 _ ________ Ave _- Park Rd A 03-41 2.600 14 In Roadway. Junction of A 200 _ _ P o 7R~_______________Pr Rd 03-44 2420 15 In Roadway A 180 03-47 2,320 16 In Roadway (Ashfray) A 100 03-51 2,000 17 In Roadway. Junction or-, __Ave A 320 03-53 950 18 In Roadway Junction of A 4050 -- Ave _ _ Ave 03-56 630 19 In Roadway. Junction of A 320 ________ Rd. Ave __ 03-59 500 20 In Roadway, Junction or A 130 =_____________ _ = =Rd 04-01 NIL 21 At Meter Posfiton A __= ____ ___ ____ ___ ____ ___ Rd 22 World Bank-30505 12 Figure 5C-10 WASTE DETECTION SUMMARY DistrIct Original Orginal Min. Original Min. Revised Revised Min. Revised Min. Reservoir W.. or Property Night Flow Night Flow Original Property Night Flow Night Flow Revised Step lest Zone No: Zone Location Count (litres) Date (1/p/h) Classification Count (litres) Date (i/p/h) Clrssitication Date Remorks 29_ 30 3 A 32 33 34 35 36 37 38 39 '0 40 41 42 43 44 45 46 47 48 49 50 52 53 World Bank-30505:15 5-31 Annex 5D. Typical Bylaws for ttie Prevention of Waste or Misuse of Water* The following bylaws are for use in the U.K. and are suitable when the plumbing system is indirect; that is, the majority or possibly the whole of the supply is delivered to a private storage tank. When the plumbing is direct and there is no private storage tank these bylaws would have to be modified accordingly. BYELAWS mdc tder Authority for prventing waste, undue consumption, mi_, at contanuaaZion of water suppl;ed by themL ujwetatin 1. In these byelaws, unless the context otherwise requires- Den _bal.W means any Boat-operated valve for controlins the inflow of water to a cistern; "3igjgj Sedud means a andard or specification issued by the Bntish Slandards Institution. " ''Cdsh 0d Cad *of Ptactm',R- a code of practice isued and avilable as aforeaid. "buM( mea=n any structure (including a floating structure) whether of a penranent character or not, and whether movable or imovablle, and without prejudice to the generality of the foregoing, indudes any caravan, vessel, boat or houseboat; a43y in rebltion to a storage cistern means the capacity of the cistern measured up to the highest level the water can reach when the bailvalve or other device for controlling the inflow of water is fitted or adjusted in the manner required by paragraph (d) of byelaw 40, or, where paragraph (e) of byclaw 41 appLies, by that paragraph; Zckwd wcwiZk means any system of pipes and other water fittings through which water circulates and from which water is not drawn for use, and includes any vent pipe fitted thereto but not the feed cistem or the cold feed pipe; naterial" means any material which is highly resistant to any corrosive action to which it is likely to be subjectel in the circunrtances in which it is used; 'cyZ mum a cylindrical closed vessel capable of containing water under presswre greater than atmospheric pressure; Mdoubt pipe" means any pipe (other than an overflow pipe or a flu&hing pipe) conveyinig water from a storage cistern, or from a hot water apparatus supplied from a feed cistem, and under presse from that ciatem; led cima" means any saorage custern used for supplying cold water to a hot water apparatus, cylinder or tank; 5-32 'overflowing Leel" in relation to a warning or other overflow pipe of a cistemn, means the lowest level at which water can flow into that pipe from that cistern; "servie pipe" means so much of any pipe for supplying water from a main to any premises as is subject to water pressure from that main, or would be so subject but for the closing of some stopvaive; "sopvalve" means any device (including a stopcock and stop tap), other than a draw-off tap, for stopping at wil the flow of water in a pipe; "storag cistern" means any cistern, other than a flushing cistern, having a free water surface under atmospheric pressure, but does not include a drinking-trough or drinking-bowl for animals, including poultry; 'tank' means a non-cylindrical closed vessel capable of containing water under pressure greater than atmospheric pressure; the undertakers" means the Authority; "warning pipe" means an overflow pipe so tixed that its outlet, whethef inside or outside a building, is in a conspicuous position where the discharge of any water therefrom can be readily seen; and "water fittings" includes pipes (other than mains), taps, cocks, valves, ferrules, meters, cistems, baths, waterclosets, soil pans and other similar apparatus used in connection with the supply and use of water. Complines 2. (1) Any requirement in these byclaws that a water fitting with Britb shall comply with a British Standard shall- Stadardan (a) be construed as requiring compliance with that Standard only in so far as the Standard relates to the size, nature, materials, strength and workmanship of that fitting; and (b) be deemed to be satisfied, notwithstanding that the fitting does not comply with that Standard in so far a it relates to those matters, if the fitting is not less efficient and suitable in relation to the purposes for which these byelaws are made than a fitting which does comply with that Standard in so far as it relates to those matters. (2) Where any requirement of any such Standard relating to any of those matters conflicts with a specific requirement of these byelaws the latter requirement shall prevail. AppUcation Applcatio 3. No person sha2l- of byelaw_ (a) use a water fitting for the purpose of conveying or receiving water supplied by the undertakers, or cause or permit a water fitting to be used, or to remain arranged or 5-33 connected so that it can be used, for that purpose, in contravention of a provision of any of the following bye- laws, that is to say, byelaws 7, 7A, 8, 10 to 12, 14 to 26 and 28 to 58A, or if the filting, or iLs mode of arrangemnnt or connection, or its situation, contravenes, or e% not in accordance with, such a provision, or if the fitting is not equipped, provided with fittings or accessories, protected or supported in accordance with or if it is equipped or provided with fittings or accessories or in contravention of, such a provision, or (b) arTange or connect a water fitting, or cause or permit a waler fitting to be arranged or connected, so that it can be used for the purpose aforesaid, in contravention of, or in a manner or in a situation which contravenes, or is not in accordance willt, a provision of any of the said byelaws, or if the fitting contravenes, or is not in accordance with, such a proviwon; or (c) alter a water fitting used for the purpose aforesaid, or cause or permit a water fitting to be altered, so that it contravenes, or is not in accordance with, a provision of any of the said byelaws, or (d) disconnect a water fitting used for the purpose aforesaid, or cause or pemiit a water fitting to be or remain disconnected, otherwise than for the purpose of repair or renewal, if, as a result of the disconnection, a provision of any of the said byelaws by virtue of which the fitting is required to remain connected will be contravened. 4. (I) None of these byelaws (with the exception of byelaws Sainp frof 7, 10(4) and 24) siall have effect so as to require any person to cease t,t - g lawfully to use, or to arrange, connect, disconnect, alter, dismantle or renew nfied any water fitting lawfully used or capable of being lawfully used on any premises immediately before the byelaw first applied in relation to those prernises, or to remove any fitting or accessory from such a fitting or to refit such a fitting in a different situation, or to provide any equipment, fitting, accessory, protection or support for a water fitting tihen used or capable of being used on any premises where the absence of [lhe required equipment, fitting, accessory, protection or support was not then unlawful, unless, by reason of the damaged, wom or otherwise unserviceable condition of the water fitting, its faulty arrangement or corineclion, its situation, equipment, fittings or accessories or the absence of the required equipment, fittings, accessories, protection or support, it causes or permits, or is likely to cause or permit, waste, undue consumption, misuse, erroneous measurement or contamination of water, or reverberation in pipes. 5-34 (2) Wlhere, for a reason mentioned in the preceding paragraph, any of these byelaws (other than byclaws 7, 10(4) and 24) has effect as mentioned in that paragraph, or where byelaw 7, 10(4) or 24 has such effect, and any work is rendered necessary by that byelaw, compliance therewith may be postponed for such time as is tessonably required for the carrying out of that work. lOee4S S. Of the following byelaws. only byelaws 6 - 8, i0 - 16, WucWuIs 23, 26, 29(3), 31, 59 and 60 shall apply in relation to any water fitting foniring part of a closed circuit. Fiting used 6. Where water is- for induztunl o# resewch (a) taken by meter ard discharged into a storage pUSpo.M cistemr, and (b) discharged into the air not less than 150mm above thle tup adse of tle cl.I,ril; slid (c) conveyed from the cistern for use for industrial or reSe;A14Lh puipobes, and it is not reasonably practicable for any one or more of the following byelaws (other than byelaws 7 and 8) to be compiied with in relation to any water fitting supplied with water from the cistern, that byelaw or those byelaws shall not apply in relation to that fitting. General provisions Unsrvceabk 7. A water fitting falling within either of the following ritn an paragraphs, that is to say- mentioned in (a) any pipe, pipe fitting, draw-off tap, draining tap, the byelawa stopvalve, ballvalve, float, cistem, cylinder, tank or ilushing apparatus or any bath, wash basin, sink, soil pan or sirnilar appliance, being a fitting which is damaged, worn ot other- wise unserviceable, or connected or arranged in a faulty manner; and (b) any fitting not specified in the preceding paragraph, whiether or not damaged, worn or otherwise unserviceable, or connected or arranged in a faulty manner, shall not be used, or be or remain so connected that it can be used, if, notwithstanding that its use or connection does not contravene any of the following byelaws, it causes or permits, or is likely to cause or permit, waste, undue consumption, misuse, erroneous measurement or contamination of water supplied by the undertakers or reverberation in pipes. 5--35 7A. Water rttings of dissiilar nietals shaJl not be used, ot be Deteriotation or remain so connected that they can be used, for the purpose of thfough electrolytic conveying or receiving the same water, uniess- ctot, (a) the circuLmlstances are such that deterioration of any of the fittings through electrolytic action is not likely to ocOuTj, or (b) effective measures are taken to prevent such deterioration. 8. (1) No pipe or cisterin used for conveying or receiving Connections water supplied by the undeiitakers shall convey or receive, or be or whkich could reniain so conniected that it can convey or receive, water not casie cn of supplied by the undertakers: mains water Provided thiat wlhere Othe water supplicd by the under- takers lo ally clitleti is discliarged into the air not less than 150mimil above the top edge therreol, this paragrapih sihall not apply to thit cistern or to any pipe conveying, or any cisterni receiving water thterefromn. (2) In thfe preceding paragraphi, "water not supplied by the undertakers" includes, and "water supplied by the undertakers" does niot include, water supplied by tihe undertakers which hias been used. (3) No pipe or cistern used for conveying or receiving potable water shall convey or receive, or be or renmain so connected that it can convey or receive, non-potable water; Provided that where the potable water supplied to any cistern is discharged into the air not less than I S0iinu above the top edge thereof, this paragraplh shaU not apply to thtat cistern or to any pipe conveying, or any cistern receiving, water therefron. (4) In the preceding paragraph "non-potable water" macniis water supplied by the unideriakers for non-domestic purposes only, arid as being unlit for driniking of culinary purposes, and "potable water" nmeans any other watei supplied by the untdertakers. (5) No service pipe or pUmiip delivery pipe drawing water from a service pipe shiall convey water front- (a) a d istributing pipe, or (b) a pump delivery pipe drawing water eitlher fron a distributing pipe or frouia a slorage cistern. (6) No pumnp or otlier means of increasing pressure shiall be installed in a service pipe for the purpose of increasing pressure or rale of flow in or fronr a service pipe or a fitting or appliance connected to that pipe. 5-36 Tap. for 9. (I ) In any premises (not being premises to which section 57 drlnking of tie Factories Act 1961 applies, or premises to which section II of the Offices, Shops and Railway Premises Act 1963 applies or which, by or under that Act, are excepted or excluded from the application of that section) in which water is supplied by the undertakers for domestic purposes, the supply ben5g separately chargeable, the person for the time being entitled or authorised to do so shall provide a draw-off tap in a position convenient for drawing drinking water on a service pipe or a po'mp delivery pipe drawmig water from a service pipe: Provided that where, by reason of the height at which the water is required to be delivered or of some other circumstance, it is not reasonably practicable to provide in the premises a service pipe or a pump delivery pipe drawing water from a service pipe, the lap may be provided on a pump delivery pipe or distributing pipe diawing water exclusively from a storage cistern which- (a) is a closed vessel having a tightly fitting access cover bolted or screwed in position; (b) is properly maintained and, where necessary, suitably lined or coated to preserve the potability of the water; (c) has an air inlet and an overflow pipe or pipes all suitably screened; (d) is, where necessary, insulated against heat; and (e) is supplied exclusively from a service pipe, or from a punip delivery pipe drawing water either from a service pipe or from a storage cistern which is a closed vessel equipped, maintained and supplied as aforesaid. (2) This byelaw shall not have effect so as to require any person to provide a tap on a pope which was in any premises immediately before this byelaw first applied in relation to those premises if the omission to provide the tap on the pipe was not then unlawl'ul. (3) No pipe on which a tap is provided in compliance with this byelaw shall be so placed that the water in the pipe is likely to become warn before reaching the tap. Protection 10. ( I ) Every water fitting, whether inside or outside a building, tfom shall be so placed as to reduce to the greatest extent which is Sinaea from reasonably practicable the risk of damage to it from frost. (2) Every water fitting (other than an overflow pipe) which, notwithstanding compliance with the preceding paragraph, is likely to suffer darnage from frost shaU be effectively protected from such damlage. 5-37 (3) The requirements of the preceding paragraphs shall be deemed to be satistied if the location, protection and insulation of the fittings are in accotdance with the recommendations in that behalf in Sections thres, four and live of the British Standard Code of Practice CP 99: 1972 "Frost precautions for watet rvices". (4) In every building and tn every part of a bUUding the supply to which is separately chargeable the water fittings on the downstreanm side of each stopvalve required by byelaw 24 shall be so arranged that they can be drained to prevent damage to them from frost, and shall be fitted with such draining taps (if any) as may be reasonably necessary for that purpose. I. Every water fitting, whether inside or outside a building, Protection which is so placed as to be liable to damage from some cause other ffom dsmage than frost shall be effectively protected from such damage. from otber 12. Every water fitting in, on or under a building shall be so Acceuibity placed as to be readily accessible for purposes of examination, repair, o ntsinp replacement and operation, except where compliance with this requirenient is not reasonably practicable or is inconsistent with the provisions of byelaw IO or II 1 Provided that this byclaw shall not prevent- (a) the enclosing of any pipe or any fitting thereon in a chase or duct if the pipe and fitting are reasonably accessible for such purposes; or (bi) the embedding of any pipe or pipe fitting in the fabric of a building so far as may be necessary for the efficient operation of any system of space heating. 13. If a water fitting is disconnected and is not within 28 days Disconnection reconnected or replaced, the person supplied with water by the of disused undertakers shall disconnect any pipe or part of a pipe which conveyed " u water supplied by theni to that fitting and is not required to convey such water to any other fitting. Fipes 14. Every pipe shall be adequately supported and shall be so Support of arranged as to avoid any air lock or reverberation. pip IS. Every pipe laid in the ground shall, unless it is under a Depth of building of a permanent character, at no time be less than 750mm pips lid nor more than 1.35m below tIhe surface of the ground, measured underpound from the top of the pipe to the ground surface. Provided that if it is not reasonably practicable for a pipe or some part of a pipe to be not less than 750rmi below the surface of the ground, that pipe or part shall be at the greatest depth below 5-3S the surface of the ground that is reasonably practicable, and in every sucih case the pipe shall be given adequate waterproof insulation and protectioni agaLInst damiage fromn causes other than freezing. Protection of 16. (1) No pipe or pipe fitting shaUl be laid, installed or allowed pipes from to remn;ain in or on the ground unless it is either of a corrosion- corTotkon and contact with r,,tiiith) itintatl2 ur affestively pmutoted from xtmereial corrosion. contimliiaing (2) No pipe shall pass into or through any ashpit, manure pit, sewer, drain, cesspool, refuse chute or any manhole connected therewitt. (3) No pipe shall be laid, installed or allowed to remain in or on any foul soid or otiier substance which could cause contamination of the water in the pipe unless it is impracticable for the pipe to be elsewhere and all necessary measures are taken to avoid any risk of contamination of the water in the pipe. (4) No pipe made of any material susceptible to permeation by any gas or otler substance which could cause contamination of the water in tihe pipe shall be laid, installed or allowed to remain in a position wliere such permeation could reasonably be expected to occu r. Deztindclcstion 16A. Every water fitting and every component of a water fitting on any pipe below ground which may be in contact with water shall be resistant or inmmune to dezincification. Provided that this requiremnenit shall not apply to any fitting or component of any litting in a closed circuit. Pipes of lead. 17. No service pipe, pump delivery pipe or distributing pipe, or wrought Vo pipe fitting connected to any such pipe, shali be of lead, wrouglt iron to be used or of steel other than stainless steel unless: - eithier (i) the circumstances are such that contamination of water or deterioration of any such pipe or pipe fitting is unlikely to occur; or (u) effective measures are taken to prevent such contanitnation of water or deterioration ol such pipe or pipe fitting. Provided that this byelaw shall not prohibit the use of a pipe or pipe fitting of wrought uon or steel which: - (a)(i) forms part of a fire sprinkler or other fire fighting installation from which water is drawn only for fire fighting purposes; and (n) is kept charged with water through an automatic aiarmn valve or is charged with water only when fire oCcu rs; or 5-39 (b) is used in connection with the supply of water for the purpose of budding, demolition or constructional work while the work is in progrcss, or tor any other temporary purpose during a period nol exceeding one month or such longer period not exceeding three months as the under- takeri may approve in any particular ca". 18. (I) I'very service pipe, puimip delivery pipe and distributing Pipes of pipe of cast iroi shall be capable of withstanding a hydraulic test cut iron pressure of not less than double the pressure to whichi the pipe will be liable to be subjected untder working conditions. (2) Every such pipe and every pipe fitting on any such pipe shall comply with Britishi Standard 4622: 1970, "Grey iron pipes and fittings" or with British Standard 4772: 1980, Specification for ductile iron pipes and liitiings (3) Where pipes are laid in ground, ile surface of which is not suitable ior heavy tralltic loads, such pipes shall be not less than Class C ol the said Bliitish Standards 1211: 1958 and 2035: 1966 provided that ithis byclaw shall not apply to a pipe laid in ground unlikely to be subjected to vehicular traffic. 19. ( I ) Every service pipe or puimp delivery pipe of steel (other Pipes of than a pipe ilo which paragraph (2) (a) of this byelaw applies) and steel every distribuiting pipe of steel in contact with the soil shall comply with the requieinemlis for heavy tubes in British Standard 1387: 1967, "Steel tubes and tubulars suitable for screwing to BS21 pipe threads". (2) Every steel pipe not in contact with the soil which is of either of the following descriptions (a) a service pipe or pump delivery pipe which- (i) forms part of a fire sprinkler or othier fire fighting installation from which water is drawn only for fire figlhting purposes; anid (ii) is kept clharged with water throughl an automatic alarmi valve or is charged with water only when lire occurs, ankd (b) a distributirng pipe, shall comply witli tile reluiremenits for medium tubes in the said British Standard 1387: 1967. (3) Every malleable cast iron pipe fitting connected to any steel pipe not prohibited by byclaw 17 being a service pipe or pump delivery pipe (other than a pipe to which paragraph (2) (a) of this byelaw applies) or a distributing pipe shall comply with the relevant 5-40 requirements of British Standard 143 & 1256: 1968, "Malleable cast iron and cast copper alloy screwed pipe fittings for steam, air, water, gas and oil", and every cast iron pipe fitting connected to any pipe to which paragraph (2) (a) of this byelaw applies sfhaU comply with British Standard 1641: 1950, "Cast iron pipe fittings for sprinklers and other fire protection installations", or with the relevant requirements of the said British Standard 143 & 1256: 1968. (4) Every pipe fitting of wrought iron connected to any steel pipe not prohibited by byelaw 17, being a service pipe, a pump delivery pipe or a distributing pipe shall comply with British Standard 1740: Part 1: 1971, "Wrought steel pipe fittings (screwed BSP Thread): Part 1: Metric Units" and every pipe fitting of steel connected to any such steel pipe shall comply with either the said British Standard 1740: Part 1: 1971 or with British Standard 1965: Part 1: 1963, "Butt-welding pipe fittings for pressure purposes: l'art I. Carbon Steel". (5) Every pipe and every pipe fitting to which any of the precedinig paragraphs of this byelaw applies shall be effectively protected fronm- (a) external corrosion; and (b) internal corrosion, unless it is a pipe or fitting whliciI - (i) forns par( of a ftre sprinkler or other fire fighting installation from which water is drawn only for fire fighting purposes; and (ii) is kept charged with water through an automatic alarmn valve or is charged with water only when fire occurs. (6) Reference in this byelaw to "steel" shall not include stainless steel. Pip" ot 20. (1) (a) Every service pipe, pump delivery and distributing copper pipe of copper connected by means of screw joints shall .comply with British Standard 2871: Part 2: 1972, "Copper and copper alloys. Tubes: Part 2: Tubes for general purposes", and the thread of every screw in any such joint shall comply with British Standard 61: 1969, "Thread for light gauge copper tubes and fittings". (b) Every copper alloy pipe fitting for a copper pipe having a screw thread which complies with Table I and 2 of the said British Standard 61: 1969 shall comply with British Standard 66 & 99: 1970, "Cast copper alloy pipe fittings for use with screwed copper tubes". 5-41 (2) (a) Every scrvice pipe, pump delivery pipe and distributing pipe of copper wliich is coniiected by means of capillary or conipression fittings or by silver brazing or bronze or autogenous welding and is laid in the Wound sJhUl comply with Britishi Standafd 2871: Part 1: 1971, "Copper and copper alloys. Tubes: Part 1: Copper tubes for water, gas and sanitallon,' l'able V. i(b) Every such pipe of copper which is not laid in the ground shall, if conrnected by capillary or compression fittinigs, or welding comply with the said Britishi Stanidard 2871: Part 1: 1971, provided that pipes to Table Z shall not be bent or connected other than by capillary 1-ittings or non-imanipulative type coimpression fittings. (c) Every capillary l-illing and compression fitting on any service pipe, pulipl) delivery pipe or distributing pipe coitiplying with sub-paragraph (a) or (b) of this paragraph sliall coinply witlh British Standard 864: Part 2: 1971, "Capillary and compression tube fittings of copper and copper alloy: Part 2: Metric Units'Xprovided that every compression fitting shalil be of l'ype B if the pipc on wlich it is fitted conipiles with British Standard 2871: Part 1: 1971: Table Y, and is laid in ihe ground. (d) Where any pipe comilplying with the said British Stanidard 2871 I)art I : 171 is connected by bronze welding by gas the welding sJiall comiply with British Standard 1724: 1959, "Bronze weldinig by gas". 21. (1) EIvery service pipe, pumip delivery pipe and distributing Pipes or pipe of asbestos cemient shall be capable of withstanding a hydraulic cabesito test pressure of riot less tihan double the pressure to which the pipe will be liable to be subjected under working conditions, and shaU comply witli British Standard 486: 1981, "Asbestos-cenient pressure pipes". (2) Where pipes are laid in ground, lite surface of which is not suitable for heavy traffic loads, such pipes shall be not less than Class 25 of the said Britishi Standard 486: 1981, provided that this byelaw shall not apply to pipes laid in grounid unlikely to be subjected to vehicular tiallfic. 22. (1) Every service pipe, pump delivery pipe, disiributing pipe Pipes of and pipe fitting not being a pipe or fitting of a material specifically other mentioned in byelaws 17 to 21 shiall be capable of withstanding a hydraulic test pressure of not less than double the pressure to which the pipe will be liable to be subjected under working conditions, and shall be of material the nature, thickness and strength of which is suitable in the cidcunastances in which the pipe or pipe fitting is to be used. 5-42 (2) Every such pipe of polythene shall comply with British Standard 1972: 1967, "Po:ythene pipe (type 32) for cold water services", or with British Standard 3284: 1967, "Polythene pipe (type 50) for cold water services". (3) Every such pipe of unplasticLzed polyvinyl chloride shall comply with British Standard 3505: 1968, "Unplasticized PVC pipe for cold water services". Where pipes are laid in ground, the surface of which is not suitable for heavy traffic loads, such pipes shall be not less than Class C of the said British Standard 3505: 1968 provided that this byelaw shall not apply to a pipe laid in ground unlikely to be subjected to vehicular traffic. (4) Every such pipe of stainless steel shall comply with British Standard 4127: Part 2: 1972, "Light gauge stainless steel tubes. Part 2. Metric Units". Filinp in 23. Every water fitting forming part of a closed circuit shall be closed of suitable material and capable of withstanding a hydraulic test pressure citcuiu of not lec.s than double the pressure to which the fitting will be liable to be subjected under working conditions. Tops and Valves Stopvsives 24. (1) Every pipe supplying water to a building (except a pipe on pipes conveying water from one budding to another building the supply to supplying buidings which is not separately chargeable and which is within the same curtilage) shall be fitted with a stopvalve inside and as near as is reasonably practicable to the point where it enters that building. (2) Every pipe supplying water to a part of a building which is a piart to whichi the supply is separately chargeable shall (unless the pipe passeS iltroughi that part to another such part) be fitted with a stopvalvc iiside and as niear as is reasonably practicable to the point wlicic tle pipe enters that part. (3) Where a pipe supplies water to a part of a budlding which ia a part to whiich the supply is separately chargeable and passes ilhroiigli thai part to another such part, every branch pipe connected to tihat pipe in the first-mentioned part shall be fitted withl a stop. valve as niear as is reasonably practicable to the point of connection. (4) Where a pipe supplies water to a part of a building which is a lulr to whlich thie supply is separately chargeable and passes hiouugl *rnc or imorc of such parts to another, it shall be so placed tlal, hbtorc entering tlie first of the parts, it passes through a place, wlhclethr iniside or outside the buildmig, to which the occupier of each ol tle parts has access, and in iliat place the pipe shall be fitted with a stoplvalve as neat as is reasonably practicable to the point where it eniters the building. 5-43 (5) Every pipe conveying water froni a building to another building the supply to whichi is not separately chargeable and which is within the same curtilage as, but has no direct access fromii, the first- mentioned building shall, subject to paragraph (6) of this byelaw, oc fitted withi a stopvalve inside anad as near as is reasonably practicable to tIlc pobin: wIhere it leaves lte first-mentioned bulidhaig. (6) Where it is not reasonably practicable to l'it a stopvalve inside the first-menetioned buildiing, tlhe said pipe shall be fitted with a stopvalve inside arid as inear as is reasonably practicable to the point where it enters the other building. (7) No szopvalve fitted in accordance with any of the preceding paragraphs of this byelaw shiall be a plug cock or plug valve. 25. Evety draw-otf pipe f'romil every cold water storage cistern of Stopvalveson a capacity exceedtng 18 litres slhall be l'itted with a stopvalve as niear to otutiel pites the cistern as is reasonably practicable. .;steris Proviided tilat, where suchi a draw-off' pipe is coninected directly to a lhot waler storage cistern, cylinider or taiik in such a way tihat it is not reasonably practicable to fit a siopvalve oui tliat pipe, a stopvalve shall be litted on every draw-of'f pipe froin the hlot water cistern, cylinder or tank, as niear tilereto as is reasonably practicable. 26. No draining tap shiall be buried in t[ie ground. or so placed Location of that its outlet is in danger ol' beiiig flooded. drariiiiig 27. No person shall erect or set up, or allow to rcna;iii crectetd Sta,tdpipe. or set up, a standpipe for conveyinig water supplied by [lie unideltakers which is used by t[ie occupants of Llmore tihani one huildinig, separately occupied part of' a building or enl unless tIme pipe is provided witih a non-coincussive self-closing tap anid a stopvalve. 28. (I) Every pipe supplying water to a drinkiikig Irougih or o)rinking drinking-bowl ior aimirimals, including poultry, shiall be litted withi a irnugl5s ballvalve or somle oilier not less elfective device for controlling the inflow of water, so desigined as to prevent overflow, amid every sucil ballvalve, or device shiall be ellectively protected riommm daninage, contamimmnat ion arid u nautborisewd mii tetererice lProvided that Itis pargraphl sihall nolt apply to a pipe if (a) the waier in tIne pipe flows by gravitalioni fromii a storLige cisterin, dmid (b) the troughi or bowl to which tile pipe suppflies water is placed at suchi a level as to prevent overflow. (2) No such troughi or bowl shall be supplied directly froim a service pipe or pumip delivery pipe drawinig water fronm a service pipe unless the inlet is fixed at a distanice above the top edge of tIne trough 5-44 o; bowl which - (a) is sufficient, having regard to the design of the trough or bowl and the use to be made of ii, to nrevCrit contam- ination of the water in the service pipe; and (b) is in no case less than 25mm. Draw-Off 29 (I) lvery metal bodied draw-off tap shall comply with the tap' relevant requirements of British Standard 5412: 1976, "Specification for the petformance of draw-off taps with metal bodies for water services". (2) I[very plastics bodied draw-off tap shall comply with the relevatit requiremients of British Standard 5413: 1976, "Specification for the perforinance of diaw-off taps with plastics bodies for water services". (3) Fvery draining tap shall comply with British Standard 2879: 1980, 'Draining taps (screw-down pattern)", or with the relevant requirements of British Standard 1010: Part 2: 1973, "I)raw-offl taps and stopvalves for water services (screw-down pattern). I'art 2. D)raw-off taps and above ground stopvalves". SIopVaWcs 30. (I) [very above-ground stopvalve not exceeding nominal sle 2 sJall comilply with the relevant reqtuirements of the said 13ritish Stanidard 1010: I'art 2: 1973. (2) L very undergrounid stopivalve not exceeding nominal size 2 shall comply with British Standard 5433: 1976, "Under- ground stopvalves for water services". (3) Lvery stopvalve exceeding riominal sizc 2 slhall comply wilh the relevant requirenients of British Standard 5163: 1974, "D)ouble flanged cast iron wedge gate valves for waterworks purposes". Operation of 31. LIvery stopvalve shiall be so placed that it can be rcaduly stopYalves operated by thc imeanis by whiich it is designted to be operated. Bativalves 32. (I) ivery balivalve of the piston type shall comply with IBritish Stan(dard 1212: Part 1: 1953, "Balivalves (exludirng floats): I'art 1: Piston type" and every balivalve of the diaphragm type shall comiply with either British Standard 1212: Part 2: 1970, "Bailvalves (excluding floais): Part 2: lhDaphragrn type (brass body)", or British Slanidard 1212: Part 3: 1979, "Float operated valves (excluding floats) I'art 3: D)iaphragnm type (plastics body) for cold water services (2) Fvery baUlvalve not of the piston type or the diaphragm vype shall couiiply with such of the following requirements as are relevait (a) every high pressure valve shall close against a working pressire of 14 bar, every medium pressure valve against a workinig pressure of 7 bar, and every low pressure valve 5-45 against a working pressure of 3 bar; and every high pressure, medium pressure and low pressure valve, not being a valve having an interchangeable orifice seating, shall have the letters 'iH.P.", "M.P." or "L.P."' cast or stamped or. t::e ody o the fitting, or shall be otherwisc clearly identifie_ Li a hn- mediuiri or low pressure valve, and every valve shIal, whiIe held mechanically in the closed position, be capable of withstanding a pressure of 20 bar; (b) the component parts of every valve of a nominal size not exceeding 50 mm shall be of a suitable and corrosion- resisting material and the lever shall be made of such material as will ensure that it does not bend under working conditions; and (c) every valve wholly or partly of ferrous metal of a nominal size exceeding 50 nun shall- Ii) be provided witlh a flange on its inlet suitable for a normal pressure of 16 bar complying with British Staildard 4504: Part 1: 1969, "Flanges and bolting for pipes, valves and filtings. Metric series: Part 1: Ferrous"; (ii) have all parts of ferfous metal protected against corrosion by coating, or by galvanising in accordance with British Standard 1387: 1967, "Steel tubes and tubulars suitable for screwing to BS 21 pipe threads", and (iii) have all ferrous working surfaces lined or faced withi, and the orifice seating made of, a suitable and corrosion-resistinig material. (3) Every bailvalve float shall comply with British Standard 1968: 1953, "Floats for balivaives (copper)", or with British Siandard 2456: 1973, "Floats (plastics) for balivalves for hot and cold waler", or witli the re(quiremilents of the said British Standard 1212: lPart 2: 1970 so far as they relate to floats of materials other than copper and plastics. Storage cisterns 33. Every storage cistern from which water is drawn for domestic Placing of purposes shall be so placed and equipped that the interior thereof StotSg can be readily inspected and cleansed, and no such cistern shall be cisterns so placed and equipped that the water therein is liable to contam- ination. 34. (I) Every storage cistern shall be adequately supported and, Support and if water for doinestic purposes is drawn from it, shall be suitably coveting of storage covered but not so as to be airtighit. cisterns 5-46 (2) Every cover shall effectively exclude light and shall be iigid with overlapping edges so constructed that it cannot easidy be dislodged and shall be of material which will not contaminate any condensate. Avodance 35. ( I ) No storage cistern shall be so placed that it is in danger of noodins. of being flooded. buried or sunken (2) No such cistern shall be buried or sunk in the ground cuslerni unless - (a) there is sufficient space around and beneath it for the purposes of mainteniance and the detection of leakage; aiid (b) either- (i) it is a closed vessel with a tightly fitting access cover bolted or screwed in position, and an air inlet and overflow pipe or pipes all suitably screened; or (is) its inlet pipe discharges into the air not less than 150 mill above its top edge: Provided that sub-paragraph (a) of this paragraph shall not apply in relation to a concrete cistern designed and constructed in accordance with the relevant recommendations in British Standard 5337: 1976, "Code of practice for the structural use of concrete for retaining aqueous liquids". NIatcr,als for 36. (1) Fvery storage cistern shall be watertiglit and of adequate slorage strength and shtall be constructed of galvanized iron or sieel,copper, (kstefns polytlieene, polypropylene, asbestos-cement, concrete or some other not less suitable material, but not lead. (2) Where the cistern is not made of a corrosion-resisting material it shall be effectively protected from corrosion. Irkitish 37. ( I ) Every storage cistern of mild steel shall comply with the ssautdardl re[quirements for Grade A or Grade B cisterns in British Standard .,, ,lase 417: Part 1: 1964, "Galvanized mild steel cisterns and covers, tanks and cylinders, Part 1. Imperial Units" or British Standard 417: Part 2 1973, "Galvanized mild steel cisterns and covers, ianks and cylinders. Part 2. Metric Units". Where Grade B cisterns are installed they shall be suitably protected against corrosion where necessary. (2) Every storage cistern of asbestos cement shall comply wilih Brtish Standard 2777: 1974, "Asbestos-cement cisterns". (3) Every storage cistern built up of cast iron plates shall comllly witth British Standard 1563: 1949, "Cast iron sectional tanks (rectangular)". 5-47 (4) lEvery storage ciisiern built up of pressed steel pl;hdes shall comply with British Standard 1564: 1975, "Pressed steel sectional reciangular tanks". (5) [-very storage cistern ol polythene or polypropylene s.JJ comply withl British Stanidard 4213. 1975, "Cold water storage cisterits (polyolefln or oleltin cupolysimer) aind cistern covers". 38. (1) YYhere in any hluse there is a cold water storage cistern Capacity of which is not connected to at y oilier suchI cistern its capacity sJIahl stir not be less tlian- houses (a) 115 litres, iltit is not used as a feed cistern; or (b) 230 litres, it it is used both as a feed cistern and lfor other purposes. (2) Where in any house there are two or more cold water storage cisterns connected togetler, the sum of their capacities shall be not less tihan- (a) 1 15 litres, if none of them is used as a feed cistern. or (b) 230 litres, if they are together used both as a feed cistern and for othef purposes. (3) In this byelaw "iouse" mians premilsses separately occupied as a private dwelling. 39. (I) Every pipe supply irng water t) a cold water storage cisterni StolragC sh;ll he lit ted witl a Iballivalve oi shaill have sto)ie otiler ihot less cis"S to he fitted cffective device for coritrollitig the inilow of water so designed as with to prevelit overflow: bialvalve. Provided iihat where two or imore cold water storage cisterins at the sanme level are conniiected together this paragraph sihall n